POLITECNICO DI MILANO

School of Industrial and Information Engineering

Master of Science in Mechanical Engineering

Development of a Wireless Controlled Olfactory Display for

VR/AR applications

Under Supervision of

Prof. Mario Covarrubias

M.Sc. Thesis of

Pedro Leal Fernandes

Matr. 804328

Academic Year 2014 - 2015

ii

iii

Acknowledgements

I would like to express my deepest gratitude to my supervisor, Prof. Mario Covarrubias for

his patience, guidance and sympathy throughout the whole duration of my research. A very

special thanks to Prof. Monica Bordegoni for helping me find the project that I wanted,

providing me with guidance and above all, patience to help me whenever I requested. I

would also like to thank my colleagues that have also provided with much valuable

assistance in dealing with occasional road blocks. A very special thanks to my parents and my

brother for being always there when I need them. Last but not least, I also want to thank all

my friends both in Italy and in Portugal that were a vital support throughout my master

studies at the Politecnico di Milano.

iv

Abstract

In the search for ways to improve Virtual Reality simulations, there have been significant

attempts into integrating the sense of olfaction by using Olfactory Displays. An Olfactory

Display is a device that can generate a range of scents and deliver them to one or several

subjects simultaneously. Recently, several different and innovative designs have been

presented, and although effective, they are often complex systems and cumbering for the

user.

This thesis consists on developing a simple, economic and reliable prototype that could be

further developed into a wearable device that would provide a better experience to the user

performing a certain simulation. A prototype was developed using simple components and a

new scent storage technology called Solid Fragrance Release (SFR) by the company OIKOS. To

control this device, an economic and reliable wireless control system was developed to

control the prototype at a distance, allowing the future user to wear the prototype without

having wires preventing him to move around freely.

Through a series of tests, both the olfactory display device and the control unit showed

satisfactory results. The device was able to deliver different scents effectively using an

imposed airflow by a DC fan that would carry the scented particles through a tube that

would deliver the scents close to the user’s nose. The control unit ensured a reliable wireless

connection up to a considerable distance using RF Transmitter-Receiver couple connected to

2 different Arduino boards, one of them connected to the device that receives information,

and the other that sends the information received by a computer interface.

v

Table of Contents

Acknowledgements .................................................................................................................... iii

Abstract ...................................................................................................................................... iv

Table of Contents ........................................................................................................................ v

List of Figures ............................................................................................................................ viii

List of Tables ............................................................................................................................... xi

Introduction......................................................................................................................... 1

Knowledge Review .............................................................................................................. 3

2.1 Smell Perception Mechanism ...................................................................................... 3

2.1.1 Odor Perception ................................................................................................... 3

2.1.2 The Olfactory System............................................................................................ 6

2.1.3 Reactions to Smell ................................................................................................ 8

2.2 Olfactory Display Design Fundamentals ...................................................................... 9

2.2.1 Technical Factors of Olfactory Displays ................................................................ 9

2.2.2 Scent Generation Methods ................................................................................ 10

2.2.3 Scent Delivery Methods ..................................................................................... 15

2.3 Olfactory Display Evaluation ...................................................................................... 20

2.3.1 Performance Evaluation Parameters ................................................................. 20

State of The Art ................................................................................................................. 23

3.1 Natural Convection Systems ...................................................................................... 23

3.1.1 Joseph Kaye’s MIT Research Projects - 2001 ...................................................... 23

3.1.2 iSmell - 2001 ....................................................................................................... 27

3.1.3 Aroma-Card Soundless Olfactory Display - 2009 ................................................ 28

3.2 Imposed Airflow Systems ........................................................................................... 29

3.2.1 Scent Collar - 2003 .............................................................................................. 30

vi

3.2.2 Presentation Technique of Scent to Avoid Olfactory Adaptation - 2007 ........... 31

3.3 Systems Using Vortex Rings ....................................................................................... 32

3.3.1 Methods and Apparatus for localized delivery of scented aerosols - 2002 ....... 33

3.3.2 ATR Media Information Science Laboratories and Meijo University Research -

2003 to 2011 ..................................................................................................................... 34

3.4 Systems Using Tubes .................................................................................................. 38

3.4.1 A CPU-controlled olfactometer for fMRI and electrophysiological studies of

olfaction -1999 .................................................................................................................. 38

3.4.2 Fragra: A Visual-Olfactory VR Game - 2004 ........................................................ 39

3.4.3 D.I.V.E. Firefighter training system - 2001 .......................................................... 39

3.4.4 Wearable Olfactory Display: Using Odor in Outdoor Environment - 2006 ........ 41

Problem Statement ........................................................................................................... 45

4.1 Possible Applications ................................................................................................. 46

The Olfactory Display ........................................................................................................ 49

5.1 Design Layout ............................................................................................................. 49

The Olfactory Display Prototype ....................................................................................... 50

5.2 Implementation ......................................................................................................... 52

Scent Generation ............................................................................................................... 52

Scent Selection and Delivery ............................................................................................. 55

The Control Unit ................................................................................................................ 57

6.1 Design ......................................................................................................................... 57

6.1.1 Arduino ............................................................................................................... 59

6.1.2 Processing ........................................................................................................... 60

6.1.3 Wireless System - RF Module ............................................................................. 60

6.2 Computer Connected Part ......................................................................................... 62

6.2.1 Interface.............................................................................................................. 62

6.2.2 Transmitter Board .............................................................................................. 63

6.3 Display Connected Part .............................................................................................. 64

vii

6.3.1 Receiver Board .................................................................................................... 64

Testing/Evaluation ............................................................................................................ 67

7.1 Olfactory Display Testing ........................................................................................... 67

7.2 Control Unit Testing ................................................................................................... 71

7.2.1 Power Tests ........................................................................................................ 71

7.2.2 Battery Life Testing ............................................................................................. 72

7.2.3 Wireless Communication Tests .......................................................................... 74

Conclusion and Future Research ....................................................................................... 77

References ......................................................................................................................... 79

Appendix A ................................................................................................................................ 85

Appendix B ................................................................................................................................ 87

Appendix C ................................................................................................................................ 89

viii

List of Figures

Figure 2.1 - Components of the Olfactory System - The Nasal Cavity. ...................................... 6

Figure 2.2 - Components of the Olfactory System - Brain Connection ...................................... 7

Figure 2.3 - System Architecture with two main functions: Scent Generation and Scent

Delivery ..................................................................................................................................... 10

Figure 2.4 - Natural Vaporization through wooden sticks from Zara home. ........................... 11

Figure 2.5 - Examples of devices with different layouts using Airflow-Based Vaporization .... 12

Figure 2.6 - Ultrasonic Atomization Principle. .......................................................................... 14

Figure 2.7 - Natural Convection Method. ................................................................................. 16

Figure 2.8 - Imposed Airflow. ................................................................................................... 17

Figure 2.9 - Vortex Ring Principle. ............................................................................................ 18

Figure 2.10 - Vortex Ring Generation through an Air Cannon. ................................................ 18

Figure 2.11 - Scent Delivery through Tubes. On the left a stationary device, and on the left a

wearable one. ........................................................................................................................... 19

Figure 2.12 - Odor Concentration vs. Time plot with the Temporal Responses parameters .. 22

Figure 3.1 - Spice rack at home with contact sensors .............................................................. 24

Figure 3.2 - Each scent is diffused using an assigned airbrush ................................................. 25

Figure 3.3 - Dollars & Scents display with the twin solenoids and the perfume bottles. ........ 26

Figure 3.4 - On the left, Scent Reminder with 5 different odors. On the right, the input side of

Honey I'm Home, a small and discrete black box that is comfortable to touch. ..................... 27

Figure 3.5 - iSmell prototypes by Digiscents ............................................................................ 27

Figure 3.6 - Transition between Gel at 25°C (on the left) and Sol at 60°C (on the right) ........ 28

Figure 3.7 - Aroma-Card with the 15 aroma-chips and the Peltier modules ........................... 29

Figure 3.8 - Smell-O-Vision was introduced in the movie Scent of Mystery, a film deliberately

created for being displayed with smell. ................................................................................... 30

Figure 3.9 - Scent Collar prototype ........................................................................................... 31

Figure 3.10 - Prototype testing ................................................................................................. 32

Figure 3.11 - Scent delivery pattern synced with breathing .................................................... 32

Figure 3.12 - Air Cannon with orifice detail. ............................................................................. 33

Figure 3.13 - Scent selector system detail inside the air cannon chamber ............................. 34

Figure 3.14 - Explanatory drawing of the layout of the system ............................................... 35

Figure 3.15 - On the left, the air cannon prototype. On the right, the camera based tracking

system. ...................................................................................................................................... 35

ix

Figure 3.16 - Third prototype of the air cannon. Notice the accordion-like section to allow for

a greater volume variation ....................................................................................................... 36

Figure 3.17 - On the left, single vortex ring hitting the user's face. On the right, the new

solution where a gentle breeze reaches the user's face .......................................................... 37

Figure 3.18 - System layout with the two vortexes colliding in front of the user. ................... 38

Figure 3.19 - On the left, camera detects hand approaching the user’s nose. On the right,

screenshot of the simulation with the user approaching a banana to his nose. ..................... 39

Figure 3.20 - D.I.V.E. Firefighter training system...................................................................... 40

Figure 3.21 - D.I.V.E. Firefighter training system in use ........................................................... 40

Figure 3.22 - Odor-Presenting Unit to be place under the nose .............................................. 41

Figure 3.23 - On the top, virtual representation of a real scent field. On the bottom, the

prototype and its components. ................................................................................................ 42

Figure 3.24 - Layout of the Direct-Injection Wearable Olfactory Display ................................ 43

Figure 5.1 - System Architecture: Layout of the system functions .......................................... 49

Figure 5.2 - 3D Model of the Olfactory Display Prototype ....................................................... 50

Figure 5.3 - OIKOS Scent Cartridges: On the left, the scented powder. In the Center, a

cartridge with the compacted powder. On the right, the cartridge used in the prototype. ... 52

Figure 5.4 - OIKOS Cube ............................................................................................................ 53

Figure 5.5 - Erosion Process ...................................................................................................... 53

Figure 5.6 - Scented Tube: Top Left, the complete scent tube. Top Right, the scent tube

without the cartridges. Bottom, Drawing view of the tube, note the section view of the tube

where air flows from the right to left. ...................................................................................... 54

Figure 5.7 - Olfactory Display Selector: On the top, highlighted selector in dark blue and the

inlet and outlet tubes highlighted in lighter blue. Bellow, side view of the selector without

the servo. .................................................................................................................................. 55

Figure 5.8 - On the left: Front View of the selector showing the tube position within an

angular range and the gaps between each tube. On the Right: A servomotor ....................... 56

Figure 6.1 - System Architecture. The Control Unit elements and their integration in the

system. ...................................................................................................................................... 57

Figure 6.2 - Computer Connected unit with the transmitter board ........................................ 58

Figure 6.3 - Device Connected unit with the board, DC motor and servomotor connected to it

.................................................................................................................................................. 58

Figure 6.4 - The Arduino Uno board ......................................................................................... 59

Figure 6.5 - Processing Software Logo ..................................................................................... 60

Figure 6.6 - RF Module with pin definitions: On the left, the Transmitter (TX) board. On the

Right, the Receiver (RX) board. ................................................................................................ 61

x

Figure 6.7 - Transmitter board circuit layout with the Pin connections of the TX module. .... 63

Figure 6.8 - The Arduino Motor Shield. .................................................................................... 64

Figure 6.9 - Receiver board circuit layout. The Pin connections of the RX module are

described in the table above. ................................................................................................... 65

Figure 7.1 - Olfactory Display with the reference points for fluid dynamic testing: P1, at the

exit of the DC fan; P2, at the exit of the selector; and P3, at the exit of a delivery tube 600

mm long. ................................................................................................................................... 67

Figure 7.2 - First 2 plots for measurements taken at the P1 and P2 points. ............................ 68

Figure 7.3 - Second couple of plots of measurements taken at P3. Above, with a tube

diameter of 10 mm; Bellow, for 5 mm diameter. .................................................................... 69

Figure 7.4 - Final plot with all the measurements taken.......................................................... 70

xi

List of Tables

Table 2.1 - Examples of varying threshold measurements of odorous substances (odorants). 5

Table 6.1 - RF Transmitter module pin definitions and purposes. ........................................... 61

Table 6.2 - RF Receiver module pin definitions and purposes. ................................................ 61

Table 6.3 - Interface keys and corresponding commands ....................................................... 62

Table 6.4 - Transmitter Arduino pin connections ..................................................................... 64

Table 7.1 - Power Consumption test results. ........................................................................... 71

Table 7.2 - Current consumption values of the Olfactory Display with the 2 batteries .......... 73

Table 7.3 - Battery life results using the equation (7.1) ........................................................... 73

Table 7.4 - Communication distance test results ..................................................................... 75

1

Introduction

Within the evolution of Virtual Reality and an increase of possible applications, new methods

to improve the realism of simulations have been introduced. One of these ideas was to

integrate the sense of smell to make the simulation more involving, but it has not been an

easy road.

Throughout the 20th century, there have been several attempts to integrate the olfactory

sense into movies, videogames and virtual reality simulations but most of them failed to

impress. The main reason is that the olfaction sense is a complex mechanism and it has its

limitations, and most developers failed to understand that. Failing to create an impact,

olfactory displays often raised the expectations too high and thus resulting in a series of

flopped commercial ventures.

In the recent years, a great variety of approaches for olfactory display designs have been

presented. Most of these developments presented good results but they are also expensive

and complex designs.

Within a project of creating a Wearable Olfactory Display (WOD), the scope of this thesis was

to create a prototype for an olfactory display using a new storage form and wireless control

unit. The system had to perform well following a simple, feasible and cost efficient design

approach. The 2 components developed are both key steps to be taken in the path to create

a WOD, a device that has to be ergonomical, effective and reliable.

The prototype tested a new scent generation method, to later be compacted into a more

portable version. The wireless control unit is a key contribution in boosting the portability of

the device. By physically separating the computer that controls the device and the device

itself, the WOD immediately becomes lighter and thus more comfortable and more versatile

which largely contributes to a nicer simulation experience.

2

This thesis begins by reviewing some fundamental aspects about the olfactory system and

the design of olfactory displays. The following chapter is the State-of-the-art, which presents

the relevant work done so far within the subject of Olfactory Displays. Chapter 4 presents

the problem statement, an introduction to what this thesis tackles within the context of

what has been done so far. Chapter 5 presents of the Olfactory Display prototype created

and Chapter 6 the development of the control unit. The following chapter 7, the testing and

respective results of the components developed in the previous 2 chapters. Lastly, Chapter 8

presents some conclusions and suggestions for future developments.

3

Knowledge Review

In this section, the background knowledge to fully understand the topic is presented. In the

first part (2.1) the mechanism of smell perception is explained, as presented by Powers [1].

Next, the basics for olfactory display engineering are described (2.2), as presented by

Nakamoto [2]. Lastly, the most relevant parameters with which olfactory displays are

evaluated and measured up against one another are explained, again based on Nakamoto’s

[2] work.

2.1 Smell Perception Mechanism

To define what are the important parameters in an Olfactory Display, it is crucial to

understand how humans perceive smell.

Vision and Audition are physical based senses for which display technologies are highly

developed. Unlike its visual and auditory counterparts, display technologies that target

olfaction are relatively underdeveloped. In addition, olfaction is a chemical sense, which

complicates the introduction of displays when compared to sensory channels based on

physical stimuli. The issue with chemical senses is its non-linearity: a change of intensity can

change qualitatively its sensation. As an example, some smells are perceived pleasant when

their light but as intensity increases, they can become distasteful.

Logically, the following section will provide an introduction to the science of smell

perception, which is paramount to understand the challenges of designing an Olfactory

Display.

2.1.1 Odor Perception

An odorant is a substance that triggers an olfactory response whereas an odor is the

sensation resulting from the stimulation of the olfactory senses. Odors play an important

role in our lives. Apart from stimulating our appetite, odors can be warning signs as several

diseases such as: gangrene, diabetes, nausea and many others have distinctive smells. In

addition, odors have can affect our moods, they are associated with memories and they can

be liked on disliked according to the associated experience of a particular person.

4

For people to feel that they are smelling something, there has to be enough concentration of

an odor to reach a certain threshold point. The odor threshold corresponds to the

concentration at which an animal panel would respond 50 percent of the time to repeated

presentations of a scent. In the case of humans, the term detection threshold is used for the

same experiment although this terminology tends to be confused.

The recognition threshold is the point where 50 percent of people can identify which smell is

being displayed. The maximum intensity that can be detected by humans is between 10 to

50 times the detection thresholds (Table 2.1) [3], which in contrast to the other senses is

low. For instances, the maximum intensity of sight is about half a million times the threshold

intensity, and the hearing that number is even higher. This low range of the olfactory system

often inhibits us from quantifying a smell’s intensity, most of the times we can only

acknowledge its presence.

The capability to perceive odors varies greatly amongst people. Over a thousand times

between the least and most sensitive individuals of an experiment panel is a common figure.

These differences vary with age, gender, smoking habits and nasal allergies. Regarding

smoking habits, adult nonsmokers show greater acuity than the common smoker. Females

usually are more sensitive to smell, something that has been made clearer through research

at the Iowa State University. Also, the olfactory nerves deteriorate as age increases; people

in their 60s only retain 38 percent of the acuity they had at time of birth.

On average, humans are able to distinguish more than 5000 different smells. However, there

are some people that have anosmia, a smell blindness that hampers them to identify one or

more specific odors regardless of their intensity. For instance, to detect natural gas leaks,

methyl mercaptan is often mixed in the composition due to its low recognition threshold

(Table 2.1) [3]. However, one in a thousand people are oblivious to this odor.

5

Table 2.1 - Examples of varying threshold measurements of odorous substances (odorants).

6

2.1.2 The Olfactory System

The sense of smell has its base on the connection between the odor stimulus and the

olfactory epithelium. The olfactory membrane covers 4 to 6 square cm in each nostril (Figure

2.1) [4]. Underneath the membrane lies a mucous layer. The olfactory cells mainly

responsible for sensing odors lie in the epithelium and the Cilia, the root shape organ that

acts as receptors for the olfactory cells, extend from the cell until beyond the mucous layer,

which increases the potential receptor area. More specifically, the specialized protein

molecules that the cilia contains are the key in the role of odor reception, and it’s an

eventual inability to synthesize a specific protein that causes anosmia.

Figure 2.1 - Components of the Olfactory System - The Nasal Cavity.

7

The olfactory cells transmit information to the olfactory bulb located at the base of the brain

(Figure 2.2) [4]. The bulb establishes the bridge between the nose fibers and the other

nerves that are connected to diverse parts of the brain.

The olfactory system has certain conditions that have to be met for it to function. For an

odor to be detected, the following conditions have to be satisfied:

1. It is necessary that scented substance has enough volatility to permeate the area.

2. The substance needs to be water-soluble to permeate the mucous layer and reach

the olfactory cells.

3. It also has to be lipid-soluble since the cilia is mainly lipid material.

4. A minimum amount of scented particles must populate the receptors for a certain

time period.

There are two main basic theories to describe the process of smell perception: the physical

and the chemical. The physical theory states that an odor being perceived depends on the

shape of the scent molecules, and that each type of molecular receiver is designed to fit a

Figure 2.2 - Components of the Olfactory System - Brain Connection

8

certain molecular shape-type of the scent. The more widely accepted chemical theory,

proposes that the odorant molecules form a chemical bond with the protein receptors that

compose the cilia. This bonding, creates a “receptor potential in the olfactory cells” that will

result in impulses being sent to the brain. The differences of detection thresholds between

individuals can be based on receptor sensitivity.

2.1.3 Reactions to Smell

Sometimes we get accustomed to an odor, like when we enter a kitchen and feel a certain

smell but then it starts to fade away. This is called odor adaptation, and what is experienced

is an increase in the detection threshold of an odor. Odor adaptation varies with odor type,

and the rate at which the detection threshold increases proportionally to the intensity of the

smell. In an extreme case of adaptation, odor fatigue occurs when there is a total adaptation

after a long and consistent exposure, and the subject becomes virtually unaware of a certain

smell.

There are some substances, whether or not they have a distinct smell, can damage the

olfactory systems and other body parts. It is known that long exposures of ammonia and

hydrogen sulfide may diminish olfactory capabilities. Pesticides also damage the olfactory

receptors, and ammonia also affects the central nervous system.

Although olfactory receptors are naturally renewed every month, exposure to harmful gases

can alter the receptors capability to regenerate. Unfortunately, the exact mechanism on how

pollutants affect the olfactory system is still reduced.

9

2.2 Olfactory Display Design Fundamentals

An Olfactory Display (OD) nowadays is referred to a device controlled by a computer that

generates scented air with the desired smell and respective intensity. The word “display” is

commonly referred to its visual counterpart, a TV or a computer screen that provides

information in the form of text or images. But instead of using visual stimuli, the OD uses

olfactory signals.

It is known that smells have a significant importance in the history of mankind, and humans

have been used for several purposes throughout the ages. For hunting in the prehistoric

years, some tribes in New Guinea used smells to build traps for animals. They would place

aromatic pieces of wood around hidden holes that when burned, released a smoke that

attracted the prey into falling inside the hole. Also, perfumes and ambient scents were used

to convey an enjoyable sensory experience to those who would be subject to it. The ancient

Egyptian civilization used aromas frequently as perfume or for therapeutically purposes. In a

more innovative approach, Cleopatra would use scented candles on her ships so that her

arrival would be preceded by a distinct alluring smell, giving her appearances a touch of

suspense [5].

The usage of olfactory displaying is still relatively unknown, although it is not a new field. In

1906, scent display was used in conjunction with cinema, even before the use of sound.

Since then, OD systems have come a long way, a topic that will be discussed further in

chapter 3. However, ODs in the computer controlled form are a recent topic.

2.2.1 Technical Factors of Olfactory Displays

Olfaction is a more complex sense compared to sound and vision, as it was explained in the

previous chapter. For a successful design to be achieve, understanding how olfaction works

is crucial.

There are several ways to develop an Olfactory Display. The device has to have 2 basic

functions - Scent generation and Scent delivery - that define the system architecture of most

devices (Figure 2.3) [2].

10

Scent Generation - is the production of scented air from the stocked odor material. It

is defined by a specific odor component and its concentration.

Scent Delivery - is the transporting of the scented air generated to the user’s

olfactory system.

These two functions aren’t always separated into different components of the system. For

example, in some systems it is logical to add the scent selection component into the system

functions. Enumerating all design types is a cumbersome task but labeling them into

separated categories isn’t easy either.

2.2.2 Scent Generation Methods

Odors are produced through essences that divide into two types: the natural and the

synthetic ones. Between the natural ones, there are animal essences that can be put into

four categories for the plant essences there are 1500 known extracts. Out of these 1500

there are around 150 on the market that are sold in the form of essential oils extracted from

flowers. Due to its high cost, rarity and difficult conservation, natural scents are usually

replaced by synthetic ones. These can come in liquid or solid form, or even as a gel and there

are around 5000 kinds from which is also possible to create more scents by mixing the

original ones.

Figure 2.3 - System Architecture with two main functions: Scent Generation and Scent Delivery

11

Vaporization Methods

There are several techniques used so far, it is difficult to name which one is best also because

they depend on the stocked form of the odor material.

Natural Vaporization

This method consists on soaking a porous material with an essence and then leave it to

vaporize naturally into the air. Just like at home when clothes are washed and they are left to

dry in the air, a nice ambient scent is left around the house. It is often used for ambience

purposes, leaving a light and pleasant smell in a room for people to enjoy for a long time.

Some examples of this method include:

A molded item with a sintered metal powder that can absorb a scent and maintain it.

A diffuser with a liquid perfume absorbing material covered by a porous material to

allow air to flow.

Some air fresheners, for example the ones that use wooden sticks soaked in a

perfume bottle that diffuses a gentle smell in the air (Figure 2.4) [6].

This is a simple and economic method but it has no odor intensity control function.

Figure 2.4 - Natural Vaporization through wooden sticks from Zara home.

12

Airflow-Based Vaporization

It works by imposing an airflow by a surface of a scented material, of any form. The airflow

can be imposed by devices such as compressors, pumps or fans (Figure 2.5) [2]. It is a very

common solution, it is simple and it can be used for both stationary and wearable devices.

This method allows for a good control of the smell intensity and it can be combined with

several scent delivery methods, making it highly versatile.

Heating

This is a method has been used for many years now, for example the burning of incense

sticks. Another example, more common, would be the aroma candles. These work by placing

liquid perfume into a small heat resistant bowl, and then heated up by a candle or a small

lamp.

For simple ambience applications it is a good solution. In addition, because there are no

mechanical components, noise and vibration are inexistent. However, it is difficult to control

the volatilized volume through heat, which is a problem when switching quickly between

scents is required.

Figure 2.5 - Examples of devices with different layouts using Airflow-Based Vaporization

13

Despite the technical difficulties, Kim et al. [7] developed a prototype that works with these

principle. They used a hydrogel that changes phase according to the temperature, and with

precise control of the temperature, the scent intensity can be controlled.

Direct Atomization

Using an ink-jet head from a printer, droplets of a scented solution can be delivered when

required. This approach allows for great control of scent delivery, both temporal and

quantitative. This method was successfully applied by Yamada et al. [8] in creating a

Wearable Olfactory Display that will be further described in chapter 3.

For this method, there are two known delivery methods. On the first case, the droplets flow

through a delivery tube or a surface, an airflow needs to be induced to atomize these

droplets into tiny particles, as in a prototype developed by Yamada et al. [8] in 2006. On the

other case, Kadowaki et al. [9] created a prototype the droplets are delivered directly into

the user’s nostrils, which means that there is no separate delivery system.

Due to its simple structure, the ink-jet head is compact and simple structured making the

device easy to miniaturize. One ink-jet head is required for each smell but it is possible to

have a device with a satisfactory scent range without compromising user comfort too much.

14

Ultrasonic Atomization

This method uses ultrasonic vibrations on a liquid essence to generate fine scented particles.

When a high frequency voltage is applied to a disk shaped piezo electric component, a

resonance is created in the direction perpendicular to the disk surface, creating ultrasonic

sound waves that will propagate through the liquid. The result is a bulging effect of the

liquid’s surface that will result in its atomization into particles, creating a considerable

amount of mist (Figure 2.6) [10].

The threshold frequency of the ultrasonic waves to create mist is of 2 MHz. When this

frequency reaches 2.5 MHz, the mist particles become very thin with a diameter of less than

3 microns. Like this, the mist will feel less humid and hover the air more easily.

Scent Switching Technology

During a video game or a movie, a certain scene starts and the correct scent needs to be

delivered, this action requires a scent switching function. This function needs to be executed

quickly, and effectively with proper isolation from the other stored essences.

Figure 2.6 - Ultrasonic Atomization Principle.

15

Several methods have been proposed so far. Some are based on mechanical switching so

that a particular odor from the stored range is selected by moving its container. For example,

in a device where the scents are stored in the slots of a revolver-like component. Others are

based on airflow control where the air is directed to the desired container. One case would

be if the scents are stored in tubes with a piece of wet cloth inside, a fan driven airflow can

be directed to the desired tube. Another would be if a pneumatic based system is

considered, where the flow would have to be directed using pressure valves. The choice of

the right scent switching method depends largely on the other components of the system.

2.2.3 Scent Delivery Methods

Once an odor is vaporized, it has to be delivered to its target. The method of delivery

depends on the several aspects according to the application: how many people have to

receive the smell simultaneously, are these people moving and how quick the transition

between smells should be. In the case where scents have to be delivered to many people, for

example a theater room, a straightforward method would be to difuse smells into the intire

area. Logically, a considerable amount of odor particles would have to be difused. On the

other hand, when the scent is to be difused to a single individual, the target area has to be as

small as possible to avoid interfering with the people nearby. For these applications,

sometimes it is necessary to have some sort of enclosure.

The duration of the presence of a smell is a function of the delivered concentration. So the

ability to control its duration will depend on how quickly this concentration can be reduced

bellow the detection threshold. For example, odors presented over a large area will take a

while to dissipate the high amount of particles. On the contrary, when targeting small areas,

the required concentration is very low so the smell will fade away immediately. This is a

great advantage when coordinating an Olfactory Display with an Audio-Visual simulation

where the odor presentation has to keep up with different scenes.

This leads us to the problem of smell removal, because a new smell cannot be delivered

while the previous one is still present. There are two approaches to solve this issue: Using a

scent elimination function or delivering a minimum amount of scent particles to the user.

In the scent elimination method, the high concentration of the present odor is reduced with

the aid of various devices, according to the amount of smell to be removed. Until moderate

amounts, suction pumps or filters can be used. For higher concentrations, ventilation

systems are usually incorporated. Ventilation would be the appropriate for large size

16

dedicated installations such as theatre rooms or amusement park rides and simulators. But

due to its complexity and cost, it wouldn’t be fit for home or office use.

With the minimum material approach, scents dissipate naturally in the air quickly. In fact,

high values of odor concentration are required, but restricted to the smallest area possible.

In this way, the particles quickly spread to the surroundings, and the scent concentration

drops below the detection threshold almost immediately. The advantage is that there is no

need for an extra piece of equipment to remove the scents, but it will require a very well

designed scent delivery system to work properly.

In the next section, the various existing methods of scent delivery are presented.

Natural Diffusion/Convection

In this method, scents are allowed to travel freely throughout the air without any induced

intervention (Figure 2.7) [2]. Scents tend to move naturally to lower concentration areas,

usually there is always a slow airflow that helps the diffusing process. Therefore, the odors

emitted are gradually dispersed throughout the room evenly.

The traditional way of enjoying ambient aromas usually are through natural convection,

giving a long lasting experience to users with subtle variations. These smells are not felt as

clearly as with other methods, but still enough to provide information about the

surroundings.

This design is quite limited in terms of spatial-temporal distribution of smells, because the

ambient air flow is mainly responsible for this. In this type of design, the distance between

the user and the device is also important, as the smell intensity is stronger closer to the

source.

Figure 2.7 - Natural Convection Method.

17

Since the ambient air flow cannot be controlled and largely influences the experience,

reducing the distance between the display-user distances will mitigate its influence.

Therefore, wearable devices can achieve higher performances than its stationary

counterparts.

Imposed Wind/Airflow

Through a fan imposed airflow, scent particles can be carried to the nose with relative

accuracy. The device can be placed at a certain distance from the user and the odors are

noticed when they reach its target (Figure 2.8) [2]. This design allows for a good control over

scent duration and good scent switching performance. Of course the speed at which these

variations are felt depend on the air flow speed and distance traveled but there is always a

limit. If the airflow is too strong to make up for a considerable distance, it can cause

discomfort to the user.

Several designs have been based on this method with very different layouts and some

examples will be presented in chapter 3.

Figure 2.8 - Imposed Airflow.

18

Vortex Rings

When distances increase, devices using imposed airflow perform poorly at some point,

because the wind can be dispersed if it travels too far. In these situations, it is better to use a

Vortex Ring (Figure 2.9) [2] to convey the odor, where the emitted scent is encapsulated in

the vortex and can reach larger distances as long the ring shape is maintained.

To launch vortex rings, an air cannon is used. An air cannon is a device that contains an open

air chamber with a circular aperture. When the chamber volume is suddenly decreased, the

air is forced out of the chamber and generates a halo-shaped vortex (Figure 2.10) [11]. By

diffusing an odor inside this chamber, scented vortex ring can be launched.

The big advantage is that the scent can be delivered at much larger distances, until the ring

collapses upon collision with the user. The scents are delivered in pulses and not

continuously, so it is not suitable for providing a long lasting odor. In addition, the scent

Figure 2.9 - Vortex Ring Principle.

Figure 2.10 - Vortex Ring Generation through an Air Cannon.

19

amount that can be delivered is defined by the size of the aperture in the air cannon, and the

ambient wind can affect its performance.

Tubes

Delivering scents through tubes is a very effective method to make odors reach the user.

There is no risk of the scents being dispersed into the surroundings and the ambient

conditions have no influence. With the appropriate scent generation and switching

procedures, this method can deliver scents both continuously and in short-term pulses,

directly under the user’s nose (Figure 2.11).

On the downside, the scent particles can adhere to the inner wall of the tubes. When this

happens, smells that were presented previously get mixed with the current smell. To solve

this issue, sometimes the scent concentrations delivered have to be very low, or more

effectively a range of tubes can be used for each smell. This problem also depends on the

form in which the odor is stocked, where liquids based odors are more prone to contaminate

the inner walls of the tubes. In addition, attaching tubes to the user’s body is somewhat

cumbersome.

Scent

Generator

Compact Scent

Generator

Tubes

Figure 2.11 - Scent Delivery through Tubes. On the left a stationary device, and on the left a wearable one.

20

2.3 Olfactory Display Evaluation

Any type of system has its performance measures. When designing olfactory displays, there

are essential factors that measure it up against the competition, and serve as goals for the

next designer. So in this section, the relevant parameters are presented.

Generally, there are two ways to evaluate olfactory displays: one is to evaluate the

performance of the device itself and the other is too evaluate the effect on users. The focus

will be on the first approach, and the factors to determine the performance of olfactory

displays are described below. There can be more factors according to other people, but in

this case these were the ones worth considering.

2.3.1 Performance Evaluation Parameters

Maximum Ability of Atomization/Vaporization

It measures the amount of scent that can be delivered at once. When the target area is large,

this becomes more relevant since users would not be able to feel the scent clearly if the

vaporization capability is not sufficient. However, if the delivery is localized, this factor loses

importance since the necessary odor concentration to be diffused is reduced.

Number of Odor Components

Logically, the more odor options a device has to offer the better. The necessary amount

varies with the application and is not a defined number, but having more options makes the

device immediately more versatile. Some devices are able to generate new odors by

combining the base scents included, but most of the times this does not work that well and

of course it is always better to use the original. However, advances have been made in

blending scents and it is likely to become widely utilized in the future.

Dynamic Range

Basically, it is how many levels of odor intensity there are to choose from. A scent generator

with a large dynamic range can control the amount or concentration of each odor

component precisely. One good example of high dynamic range would be the ink-jet

olfactory display developed by Kadowaki et al. [9] (3.2.2) that has 256 levels of intensity.

21

Accuracy

It can also be referred as stability or repeatability, it measures the ability to keep the odor

concentration level constant, matching the desired value. This capacity depends on the type

of scent generation. For example, in scent generators using airflow-based vaporization, the

concentration of odor material depends on the velocity of the imposed airflow. In these

cases, making sure the airflow characteristics are constant is paramount to ensure a constant

odor intensity.

Crosstalk

Crosstalk refers to the phenomenon that unintended odor component(s) are mixed in the

desired combination of odor components. This can happen more commonly for example, on

tube based systems. If a tube long tube is used to deliver the scent, odors can adhere to the

walls. On the other hand, the cooling-based scent generator developed by Kim et al. [7]

proved very good in this matter, as the odors don’t are vaporized into the air directly

through heating.

Temporal Response

Temporal response measures how fast the system reacts to commands and it is one the most

important performance aspects. Parameters include delay, rising time, sustained period, and

decay time. The plot (Figure 2.12) [2] shows the general temporal aspects used to evaluate a

pulse wave, where it is assumed that the target concentration level is set to 𝐶𝑡 at t = 0 and

reset to zero after a time period. The scent concentration delivered over time cannot be

entirely controlled with scent generation, it has to be evaluated using odor sensors. Note

that, decay time is usually longer than rise time by default.

22

Efficiency

Efficiency measures how much of the odor diffused is actually consumed by the user. Better

efficiency means less wasted odor material, which means that odor supplies need to be

refilled less often. On this aspect, wearable olfactory displays perform particularly well,

followed by systems using vortex-rings.

Comfort

It is impossible to quantify this factor, but it is one the most relevant nonetheless.

Considering the previous parameters, the wearable approach offers a solid response and

good efficiency. However, one can also notice the tradeoff between these desired

performances and the cumbersome impression of wearing a device. Therefore, it is

important to select an appropriate method, depending on the application and the purpose of

presenting scents.

Figure 2.12 - Odor Concentration vs. Time plot with the Temporal Responses parameters

23

State of The Art

The Olfactory Display devices developed so far are as numerous as they are diverse, even

though it is considered to be a relatively new field. One of the main reasons for this could be

that there hasn’t been a system that had a significant commercial impact.

In this section the most relevant related works are presented, including academic researches

and industrial products. In some cases, several projects were launched by the same group of

people, because they have been researching on this topic continuously for quite some time.

These developments are organized by scent delivery method, like in the previous chapter.

And within each category, it follows a chronological order.

3.1 Natural Convection Systems

As mentioned before, the first use of scent display to enhance the “realness” of a simulation

dates back to the beginning of the XX century, and it used natural convection. The

conjunction of film and theatre started in 1906, when a Philadelphia cinema owner named

S.L. Rothafel sprayed the audience with the scent of roses during a screening of the Rose

Bowl [12].

3.1.1 Joseph Kaye’s MIT Research Projects - 2001

During his MIT Master of Science, Joseph Kaye [12] wanted to explore the possibilities of

conveying information using scent. The ultimate goal was to use olfactory as a

communication channel, in the same way as computers can play music.

Within this scope, he developed 5 prototypes that caused some impact in the field. The

importance of his work is not related to the development of technological wonders in the

device’s hardware, but the introduction of very innovative and useful ideas using smells in

common day-to-day situations. In fact, the systems are relatively simple, suggesting that the

integration of this technology in our everyday lives could be implemented sooner than

expected.

The scope of this thesis is more about ways to build olfactory displays, rather than possible

uses for the technology. Nevertheless, the sheer creativity in these ideas backed by a strong

24

knowledge about human behavior lead to the development of very interesting work. Such

ideas have shown that there might be a relevant marketability in the world of Olfactory

Displays after all.

“inStink” - February 2000

Let’s imagine the situation of being abroad, and one is missing the little things that make

home feel like home, like the smell of home cooking. The goal here is to develop a device

that brings that sensation to the user, even though home is very far.

Ideally, to recreate environments as accurate as possible, the system would have an

electronic nose at home that could recognize the smells, and communicate with the other

end which would recreate the odor to be presented.

Since this wasn’t possible, there is a spice rack at home and a device that detects when a

specific spice is being used. On the other end, a range of scents of those spices is available so

that the spice being used at home can be diffused (Figure 3.1) [12].

The two components of the system work as follows. The first part, at home, consists of a

spice rack with a contact sensor attached to each specific spice jar (Figure 3.1). A

programmable board receives the signals from these sensors. These signals are then sent to

the other end, where smells were diffused using individual airbrushes (airbrushes pic). These

airbrushes would be controlled by another programmable board that controls a valve system

fed by a CO2 tank or a small air compressor (Figure 3.2).

Figure 3.1 - Spice rack at home with contact sensors

25

In terms of getting a real experience, the device is limited to the available spices, leaving out

all other cooking smells. In terms of scent generation and delivery, both functions worked

without problems. However, it is a complex and expensive system that would not be fit to

adapt to a Wearable version. The main issues are related to the quality of the smells, and the

impossibility to recreate an olfactory experience that accurately matches the user’s memory.

“Dollars & Scents” - October 2000

The second project was a lot simpler. The idea came from a previous thesis by Wisneski [13],

who designed a pocket device to present information of the stock market to the user by

becoming hot or cold. In addition, through research Ren [14] discovered that people wish to

be aware of the state of the market but do not want to dedicate full attention to a detailed

analysis constantly. Therefore, displays can give an idea of the market without disrupting the

attention of the user.

So, the system consists of two spray bottles that are operated by two solenoid valves (Figure

3.3). The two solenoids are controlled by a single-board TINI computer, which launches the

smell of mint if the market is going up, and lemon if the market is going down. The force

required to push down the plunger of the perfume bottle demands a large solenoid valve,

which requires considerable power and room to accommodate.

Figure 3.2 - Each scent is diffused using an assigned airbrush

26

The response was generally positive, the device did not cause any discomfort and people

enjoyed it. There was also the idea of regulating the intensity according to the level of the

market rise/drop, but this would require an almost still airflow in the ambient and people

adapt to scents very quickly, which would allow only to note the smell changing [12].

Using the same technology, Kaye came up with two more devices: “Scent Reminder” (Figure

3.4), a computer controlled “olfactory schedule” that works in conjunction with Microsoft

Outlook, diffusing a specific smell to remind the subject of a specific event. And, “Honey, I’m

Home” (Figure 3.4) [12], a new concept conceived by the author and accurately described in

the following quoted piece of text: “On my girlfriend’s desk sits a small, rounded, black box.

At the back of my desk is light blue acrylic structure. When my girlfriend wants me to know

she’s thinking of me, she rests her hand on the box for a couple of seconds. A gentle, warm

scent of hazelnut wafts across my desk, without interrupting my meetings or phone calls. It’s

good to know you’re loved.”

Figure 3.3 - Dollars & Scents display with the twin solenoids and the perfume bottles.

27

3.1.2 iSmell - 2001

In 2001 Digiscents announced iSmell, a computer controlled device that could spray scents to

its user while accessing a website or opening an email. The company initially started with

working on movie clips with integrated sequences of smell in a project called “Scentracks”

[12].

After an enthusiastic article was published in Wired Magazine, the company decided to

produce a device for home use. It would be controlled by USB or serial port and it would

contain a cartridge of 128 primary odors that could be mixed to replicate a wide range of

smells [15] (Figure 3.5) [16].

Figure 3.4 - On the left, Scent Reminder with 5 different odors. On the right, the input side of Honey I'm Home, a small and discrete black box that is comfortable to touch.

Figure 3.5 - iSmell prototypes by Digiscents

28

The company announced in fall 2000 that it would start commercializing the product in the

following spring, but by April 2001 not even the developers had hardware devices. The

prototyping phase was never finished and shortly after, the company bankrupted. It became

famous because there was a hype about the announcement on the internet, but the

standard was set too high for the Digiscents to ever come through with a viable solution.

3.1.3 Aroma-Card Soundless Olfactory Display - 2009

The research group led by Kim et al. [7] presented a rather innovative approach. Usually,

electro mechanical devices are used to deliver the scents such as fans and compressors.

Therefore, most of the times these devices are noisy and vibrate to some extent, even

though there are fans that are barely noticeable. As solution, the system uses a temperature

responsive hydrogel for each scent that alters between sol and gel (Figure 3.6) [7], and so it

aroma release is accurately controlled by a “Peltier” module to control the temperature.

The scents are stored into 15 chips that are placed on top of 15 “Peltier” modules, a thermo-

electric cooling device used upside down to provide heat, and their temperature is controlled

individually by a computer. This board that includes the 15 aroma-chips and “Peltier”

modules is called the aroma-card and is located at the top of the device (Figure 3.7) [7].

Figure 3.6 - Transition between Gel at 25°C (on the left) and Sol at 60°C (on the right)

29

The results showed potential although there are some inherent disadvantages. The

responsiveness of the system is slow since it takes about 10s between the heating start and

the user noticing a smell intensity variation. Apart from the reduced noise, there is no

adhesion of scent particles to the device since the aroma-chips sit on top.

3.2 Imposed Airflow Systems

Since that first scented theatre experience in 1906, fifty years had passed until the next

development. With the television boom in the late fifties, Cinema owners were worried that

their clientele would be reduced. And so, they looked for new ways to make cinema-going a

more attractive activity.

In this context, AromaRama came out in December 1959, mixing smells with a travelogue of

China called “Behind the Great Wall” [12]. The system used Freon gas to diffuse smells using

the air conditioning ducts to the theatre rooms. It created an expectation that did not match

reality: odors seemed fake, they were strong enough to cause headaches to some of the

viewers and they weren’t removed quickly enough to keep up with the scenes changing. One

year later, Smell-o-Vision (Figure 3.8) [12] was presented and despite a more complex

system using a tube for each seat to deliver scents, it failed to impress the critics as well.

Figure 3.7 - Aroma-Card with the 15 aroma-chips and the Peltier modules

30

These earlier systems were used for large scale applications and made evident several

problems in Olfactory Displays, namely smell removal. Knowing this, several systems came

out decades later that provided far better results using the same delivery system.

3.2.1 Scent Collar - 2003

Scent Collar [17] is a wearable olfactory display created by the Institute for Creative

Technologies at the University of Southern California. With both aesthetics and functionality

in mind, the Scent Collar is a collar shaped device that was designed to be used with virtual

reality simulations.

To avoid contaminating the air in simulations, a minimum amount of odor is diffused. For

this to work the device needs to be close to the users face. Hence, the collar design. The

components were meant to be as compact as possible to maximize the scent range without

jeopardizing comfort.

Ultimately, the presented prototype (Figure 3.9) [17] houses 4 scents that were stocked as

oil-soaked wicks in individual slots and then diffused using small fans. The device is

controlled by Bluetooth in a virtual space with scent-marked zones for the user to move

around: scents are activated when the wearer enters a marked location of the virtual space.

Figure 3.8 - Smell-O-Vision was introduced in the movie Scent of Mystery, a film deliberately created for being displayed with smell.

31

The initial goal was to have a range of 10 aromas. To make this happen, the system was to be

equipped with Micro-Electro-Mechanical Systems (MEMs) but a prototype was never

presented.

3.2.2 Presentation Technique of Scent to Avoid Olfactory Adaptation -

2007

During long movie scenes, viewers cannot feel odors being diffused continuously over long

periods of time because people eventually adapt to a smell. To solve the problem, Kadowaki

et al. [9] Proposed delivering scents in small pulses.

The display uses an ink-jet olfactory diffuser by CANON that ejects droplets. On each of the

12 scent tanks there are 256 micro-holes that can be shut individually to provide a very

accurate control over the quantity delivered. After ejection, the droplets are atomized by a

fan and delivered to the user’s nose (Figure 3.10) [9].

The olfactory ejection is synchronized with users breathing pattern, delivering a pulse of

scent when a user inhales. The system times the breathing using a thermistor that senses

temperature variations over time: when the user breathes out the temperature rises. With

this feature, the scents are delivered while the user is breathing in (Figure 3.11) [9].

Figure 3.9 - Scent Collar prototype

32

The results were satisfactory, users could feel the same scent for a longer period. Although

this device is highly complex and it does not consider usability and comfort for the user.

Apart from that, it requires a complex and expensive apparatus. Nevertheless, a pulsed smell

delivery pattern can be used in several other types of stationery and wearable devices and

improve the quality of the simulation.

3.3 Systems Using Vortex Rings

The interest of these systems came about with the limited range of fans in delivering scents

over a certain distance with a considerable accuracy and wasting less amount of scent. The

Figure 3.10 - Prototype testing

Figure 3.11 - Scent delivery pattern synced with breathing

33

fact is that there aren’t that many systems out there that use vortex rings: maybe because

there is not a big enough demand of its distance capabilities, or maybe because the

apparatus is too complex to attract people researching into it. However, this method is new

when compared to the other ones and maybe be the right solution for a specific application.

3.3.1 Methods and Apparatus for localized delivery of scented

aerosols - 2002

This display developed by Carl J. Watkins [18] is the first to use a Vortex Ring. It is composed

by a box that with an orifice and it houses a speaker and multi-scent generator inside. The

size and distance reached by the ring can be adjusted by changing the orifice diameter

(Figure 3.12) [18] and the frequency of the pulse.

Before a vortex ring is launched, the inner chamber needs to be filled with scented air. The

scent generator contains several (number not specified) recipients of liquid based aromas

that are connected to an electro-pneumatic system, using a valve system and a pressure line

to select the desired scent to permeate the air inside the chamber (Figure 3.13) [18]. There

are no reported tests or any results to prove the system was successful, but the concept was

indeed used for developing other devices.

Figure 3.12 - Air Cannon with orifice detail.

34

3.3.2 ATR Media Information Science Laboratories and Meijo

University Research - 2003 to 2011

The title of this section seems a bit odd but there is a reason for it. Throughout 8 years, a

group of Japanese scientists started developing an olfactory display using an air cannon.

Along the way they kept adding new features to it, new people appeared to help and 4

articles were published. So this section, will be divided into 4 subsections for each new

contribution to the system.

An Unencumbering, Localized Olfactory Display - 2003

The project was kicked off by Yanagida et al. [19] with the goal of making an Olfactory

Display to be included in the next-generation virtual reality systems. In the attempt of

making a device that performs well and promotes comfort by avoiding a wearable system,

the group proposes an “unencumbering” device (Figure 3.14) [19], using an air cannon to

deliver the scents with precision.

An air cannon was developed, keeping in mind that the scent emission has to be launched to

a precise area. Accuracy was a top priority. This way, wasted odor material is reduced and

different scents can be displayed to multiple users. To make sure the air cannon is aiming at

the right place, a face tracking system was developed to include in the system. The tracking

Figure 3.13 - Scent selector system detail inside the air cannon chamber

35

system used a camera and it worked by tracking the eyes of the user [20], and then it would

track the nose in its vicinity, since it usually has a highlight pattern that can be detected by

the camera (Figure 3.15) [21].

The air cannon in these experiments was made out of cardboard and acrylic plastic with an

output hole in it (Figure 3.14) [19]. When one of the walls was pushed, a vortex ring is

launched at 1 meter per second. In this preliminary stage, there is only one smell available.

The scent is delivered by a tube and sprayed inside the box just before the ring is launched.

By doing so, several rings with different odors can be launched to multiple users.

Figure 3.14 - Explanatory drawing of the layout of the system

Figure 3.15 - On the left, the air cannon prototype. On the right, the camera based tracking system.

36

The results were promising: out of 59 launch attempts there were 11 failed, and 9 missed

hits. Out the remaining 39 that reached the targets, there was a smell detection rate of 72%

(28 out of 39). From this point, the focus was on improving the design of the air cannon and

the inclusion of a scent switching function.

Projection-Based Olfactory Display with Nose Tracking - 2004

On this second step of the research, which is actually the 3rd prototype developed by the

group, a scent switching function was integrated by Yanagida et al. [21]. The issue was that

by ejecting the scent inside the air cannon chamber, the smell would not clean properly with

the vortex ring launch. Cleaning the scent from the chamber would be difficult, and also the

air cannon could be improved, so a new one was built.

The new air cannon sits on a 2-DOF custom made platform that is connected to the camera,

like in the previous model. The volume variation was increased in this model by including an

accordion-like chamber (Figure 3.16) [21] that is controlled by a stepping motor. To solve the

problem of scent switching, a short cylinder with the same diameter as the aperture of the

air cannon was added and it included mechanical shutters at both ends. On the side faces of

the cylinder, 4 holes were made for scent delivery and 1 hole for cleaning. Attached to these

holes are tubes that are managed by an electro-pneumatic system controlled by an outside

computer.

Figure 3.16 - Third prototype of the air cannon. Notice the accordion-like section to allow for a greater volume variation

37

SpotScents: A Novel Method of Natural Scent Delivery Using Multiple Scent Projectors - 2006

Having satisfying the basic requirements, the research focused on fine tuning the experience.

With the current system, users felt a strong unnatural airflow when the vortex ring would hit

their faces. To reduce this effect, Nakaizumi et al. [22] used two air cannons to launch vortex

rings that collide close to the users face. Upon collision, the rings break and smell is

distributed around a small area or “spot”. With this configuration users felt a gentle scented

breeze, and the experience became a lot more pleasant (Figure 3.17) [22].

The results showed that there was in fact a reduced airflow around the target points.

However, the range of the scent delivery is reduced compared to just using a single air

cannon. It proved hard to ensure an accurate collision above 1100 mm.

Localized Scent Presentation to a Walking Person by Using Scent Projectors - 2011

The system could detect the user’s face to deliver the smell, but they could not be applied to

a moving person. In this experiment, Murai et al. [23] included a time-of-flight base range

imaging camera to track a user entering the area of interest. In the experiment, it is assumed

that the user is walking in a straight line at constant speed, in an attempt to recreate an aisle

of a super market.

Figure 3.17 - On the left, single vortex ring hitting the user's face. On the right, the new solution where a gentle breeze reaches the user's face

38

The system layout is shown in the (Figure 3.18) [23], two air cannons are set 1m apart. The

system calculates the user’s velocity and aims two vortex rings to collide 50 cm in front of

the users estimated position. As a system that has an inherently small margin for error, the

results were not very satisfactory. Only 66% of the experiments were successful in making

the user perceive the smell.

3.4 Systems Using Tubes

3.4.1 A CPU-controlled olfactometer for fMRI and electrophysiological

studies of olfaction -1999

This is a design for a reliable and economical olfactory display. Using an electro pneumatic

system, with pressure lines and solenoid valves, a system was created by Lorig et al. [24] that

works seamlessly, ensuring quick rise times, effective scent switching and cleaning.

The use of electro pneumatic circuits in Olfactory Displays has been proved to be very

effective. Joining that to tube delivery, and an accurate localized delivery is guaranteed.

However, this system is as good as it gets the way it is. There can be issues with odor

adhesion to the tubes and in this configuration, this concept could not be further developed

into a compact/portable device.

This system is a great introduction to exemplify the potential of tubes. Yet, to get a wearable

device or a more compact solution, different scent generation and delivery technologies

need to be used. The following projects are remarkable efforts in making wearable devices.

Figure 3.18 - System layout with the two vortexes colliding in front of the user.

39

3.4.2 Fragra: A Visual-Olfactory VR Game - 2004

By understanding the relationship between olfaction and vision, our lives can be enhanced

by communicating through the olfactory channel: cooking programs that allows us to smell

the dishes on demand, a new type of restaurant menu which we can see and also smell. With

this in mind, Mochizuki et al. [25] developed a visual-olfactory display device along with a

game “Fragra” to research the interaction between these two scents.

The idea was to simulate the action of a person grabbing a piece of food or flower, to smell it

by approaching it to the nose. Scented air is delivered through tubes to the user’s hand, on

the contrary to the face like most devices. The system detects the hand approaching the

user’s nose, by using a camera and markers on the user’s hand (Figure 3.19 left) [25], and

then it delivers the smell. The scent is selected using a PC controlled solenoid bulb according

to the object displayed in the game (Figure 3.19 right) [25].

3.4.3 D.I.V.E. Firefighter training system - 2001

The Deep Immersion Virtual Environment Laboratory at the Southwest Research Institute in

1992 started developing a firefighter training system: an olfactory display designed to train

firefighters delivering odors in fire simulations.

Lead by John Cater [12] [26] [27], the team developed a backpack mounted device (Figure

3.20) [27], with scents delivered through the oxygen mask that is used in the regular

firefighter equipment. The system proved itself from the beginning, and so developments

came with the years.

Figure 3.19 - On the left, camera detects hand approaching the user’s nose. On the right, screenshot of the simulation with the user approaching a banana to his nose.

40

The odor quality was improved and the scent selection range as well. The final design in

2001, used fluid essential oil wicks developed in conjunction with Fragrance Technologies of

Windermere, Florida that produced the best results achieved so far. The most impressive

feature of the device, is that the olfactory output that ranges from a barely noticeable odor

to an “unbearable stench that makes you want to rip the mask off” (Figure 3.21) [27].

Figure 3.20 - D.I.V.E. Firefighter training system.

Figure 3.21 - D.I.V.E. Firefighter training system in use

41

3.4.4 Wearable Olfactory Display: Using Odor in Outdoor Environment

- 2006

Besides creating a device, in this work by Yamada et al. [8] the goal of this was also to create

a virtual space where scents vary in type and intensity according to the user’s position. So, in

this research, a wearable olfactory display was created that delivers odors in the gaseous

state and uses tubes to deliver it to the user’s nose.

The device’s odor generating and control units are “worn” as a backpack. The tubes come

from the backpack and arrive at a Head Mounted device. The odors arrive from several tubes

at a larger cylinder that is located under the user’s nose, and they are delivered through tiny

side holes of this cylinder (Figure 3.22) [8]. The system also includes a tag reader, to read

markers on the floor that enable to detect the odor sources and create the odor field (Figure

3.23) [8].

Figure 3.22 - Odor-Presenting Unit to be place under the nose

42

The odor delivery increases strength when the user reaches any of the sources. The airflow is

generated by a DC pump. Since the system has a tube for each odor and one tube for clean

air, the strength of the presented smell is controlled by managing the proportion between

non-odor airflow and odor airflow.

In addition, the research group developed another prototype but with a different scent

delivery method that wasn’t referred beforehand - Direct Injection. This system can be

described by not having a delivery method per se, scents are generated at the user’s nose

and so they don’t need to be “carried”. This is achieved by placing several ink-jet heads, like

the ones seen before, close the user’s nose. Apart from that, the system uses a breath

detection unit that tells the control unit when to deliver the scents, to avoid a continuous

pattern. To have successfully integrated these components into a wearable device (Figure

3.24) [8] is a remarkable achievement, even though user comfort is extremely compromised.

Figure 3.23 - On the top, virtual representation of a real scent field. On the bottom, the prototype and its components.

43

Both prototypes performed well in the testing phases. With the tube delivery prototype, it

was possible to identify odor the odor sources with varying intensities. The direct injection

prototype proved to be superior in its operation life and stability in presenting odor strength

over the first prototype.

Figure 3.24 - Layout of the Direct-Injection Wearable Olfactory Display

44

45

Problem Statement

The goal of this project was to develop a wearable olfactory display for Augmented Reality

and Virtual Reality purposes. More specifically, the scope of the project was to develop a

working prototype of an Olfactory Display that would serve as the base for further

developments, and to create a wireless connection system that would enable the device to

be controlled at a distance, a key element in making the device more compact and it´s use

more comfortable.

As a key element on the scope of this project, the creation of a functioning wireless system is

just not enough. To make it viable, it must be within a certain cost range without

compromising on reliability. In addition, the power source would have to be portable as well.

Addressing these issues was one of the goals of this project, even though they seem to be

frequently overlooked.

The novel feature of this device is the scent storage form. Most devices use either liquid or

gas based aromas, this device uses cartridges of compacted scented powder produced by the

company OIKOS fragrances [28]. The company makes scent diffusers for ambient uses using

an innovative patented method of scent storage and delivery called SFR (Solid Fragrance

Release). It is a simple principle: by applying a gentle airflow on the solid scent cartridges

that release particles in the air. This technology has some advantages compared to the

traditional ones that open up new possibilities in Olfactory Display design.

Often, Olfactory Displays are complex systems. There seems to be a generalized focus on

making perfect smell delivery systems, and usually researchers focus on innovative and

complicated technological solutions. Mostly, the results of these developments are

satisfactory but usability and practicality are usually not taken into account. Despite

interesting, the concepts presented are too complex to ever be used in the real world and so

they never leave the laboratory.

As a result, there is very little commercial interest in the topic. Devices exist, they can deliver

smells, but simulations lack appeal and a sense of reality. It can be conclude that the majority

of the research done, lacks a commercial drive. Developers often neglect ergonomical

aspects such as comfort, wearability and design feasibility in their developments.

46

On the contrary, the design proposed here aims for simplicity, low cost and user-friendliness,

without disregarding a possible marketability of the end product. The OIKOS scent cartridges

due to its easy scent delivery mechanism, are more than adequate for such an application.

More concretely, the objective is to assure the main functions of an olfactory display - scent

generation, selection and delivery - and a wireless control unit with a simple interface. In

addition, some guidelines were to be followed:

1. In a later stage, the end product is a wearable device. In other words, the end

product has to be compact, simple, and comfortable, so whatever is developed in this

early stage needs to be convertible to suit the final design requirements.

2. Costs are to be kept low, which means the device can only include simple prototyping

mechanisms that are simple, reliable and economic.

4.1 Possible Applications

To fully understand the importance of any innovation, it is essential to know what it is for.

There are many approaches into designing Olfactory Displays, but the right choice depends

on its future application. In this part, some possible uses are presented for this technology.

Cinema and Movies

As it was presented before, the first goal of an olfactory display was to enhance the

experience of watching a movie by coordinating scents with the apropriate scenes. The first

ever application was in 1906 when S.L. Rothafel sprayed the audience with a rose scent

during the screening of the Rose Bowl, the most important event in American college

football. After that, examples include AromaRama in 1959 and Smell-O-Vision one year after

by delivering different a range of scents to the audience. Another way to deliver scents was

to use scratch and sniff cards that were first used in the 1980s until the early 2000s with the

Spy Kids franchise [12]. Initially this created quite a buzz but the actual experience could

never match the expectation, so it never really became popular.

Virtual Reality and Gaming

More recently, the rising popularity of simulations opened a window for the use of olfactory

displays. For amusement purposes, some video games included scents. Examples include

“Fragra: A Visual-Olfactory VR Game” [25], the iSmell device by Digiscents [12] that never

cleared the prototyping phase and the device developed by Nakamoto et al. [29] that

47

attempted to create a cooking game that would release smells of each added ingredient to a

virtual sauce pan. With a different purpose but within the topic of simulations, the D.I.V.E.

Firefighter training system [12] delivered smells using the mask in the American standard

firefighter equipment to simulate fire situations for training. Also, it was noted that surgery

simulations in medical schools used to train future surgeons lacked a crucial olfaction

component in their simulations. Through smells, doctors can detect infections amongst other

things. However, creating accurate smells that resemble the real ones is too difficult, so such

a device was never developed.

Medical

Aromas can have a therapeutic effect on people and many ancient civilizations were aware

of this. The use of essential oils for therapeutic, spiritual, hygienic and ritualistic purposes

goes back to a number of ancient civilizations including the Chinese, Indians, Egyptians,

Greeks, and Romans who used them in cosmetics, perfumes and drugs [30] thus inventing

Aromatherapy. In addition, some olfactory displays may include an air purification function.

Through a chemical reaction between the scented air and the ambient air, some undesirable

elements may be eliminated.

Memory Triggering

It is known that smells have memories associated to them and this idea has been explored to

develop innovative concepts. The Jorvik Viking Museum in York, England for example, used

odors that successfully increased people’s ability to remember the information presented at

the exhibit [12]. Another example is the work developed at the University of Glasgow by

Stephen Brewster et al. [31] that tried to adapt the memory triggering effect to help browse

to large digital photo collections, attempting to assign smell tags to specific pictures that

would help the user to recall them while looking for them. A side effect of recalling

memories through smells is the triggering of associated emotions. One idea was to enable

couples to send scent signals to one another while they would be apart, a concept explored

by Joseph Kaye’s “Honey, I’m Home” [12] at MIT.

48

49

The Olfactory Display

5.1 Design Layout

In this part, the system architecture of the system is presented (Figure 5.1). That is, the

overall layout of the system to present a general perspective to the reader. A more detailed

explanation of each element will be presented subsequently.

As mentioned before, the system is composed by the Olfactory Display device and the

Control Unit. The Olfactory Display is, to put it briefly, a tube-based system where air flows

through. Within the Olfactory Display, there are 3 key functions:

Figure 5.1 - System Architecture: Layout of the system functions

50

Scent Generation - The odor is released by imposing an airflow on the scent

cartridges place inside small PVC tubes. The airflow is created by a Direct Current (DC)

powered fan that releases odor particles through an erosion process.

Scent Selection - The different scent types are placed inside tubes, each tube

corresponds to a specific scent, except for one of that is intentionally empty for the

cleaning function. The tubes are fixed to a rotating cylinder, similar to a revolver gun.

Using a servo motor, the cylinder rotates to the desired position to select a specific

smell to be delivered.

Scent Delivery - After passing through one of the tubes of the selector cylinder, the

scent goes through a flexible plastic tube to be delivered close to the user’s nose.

The Olfactory Display Prototype

Figure 5.2 - 3D Model of the Olfactory Display Prototype

51

In the picture above (Figure 5.2), the model of the device is presented. This is a

representation of the final desktop version developed. It delivers smells accurately, without

any noticeable crosstalk and it quick at responding to the user’s commands.

Starting upstream, the big component in light brown is the DC centrifugal fan that draws air

and delivers it to the system. A DC motor is a device that is widely used in all sorts of

applications, it works by creating an electromagnetic field creating a torque on the rotor. The

higher the input voltage is, the more current will be drawn and consequently more power

and torque. There are 2 types of DC motors: brushed and brushless. In our model we used a

brushed dc motor, that is cheaper and simpler in its design but its longevity is limited and it is

noisier. The brushless is quieter, more reliable, more efficient, and lasts longer. But it is also

more expensive. For prototyping purposes the DC motor was not a major concern as long as

it fulfilled the task of delivering scents, but this leaves room for improvement, and a

brushless dc motor is most likely the better option [32].

At the outlet of the fan is the inlet of the scent selector in light grey, which uses a funneled

shape tube to reduce the airflow cross section to fit the smaller selector tubes. In light blue,

the revolver cylinder shaped selector with its 4 tubes that is driven by the servo motor, the

small component in light brown. Lastly, the light red tube at the end of the device is the

outlet that will lead to the transparent plastic tube that ends at the user’s nose.

Before arriving to the final result, there were some failed attempts in middle. A detailed

description of these early prototypes won’t be presented, but they made evident what

would be the key issues that prevented the system to work properly. More concretely, it

showed that a fluid dynamic efficient design was paramount. Pressure losses had to be

minimized to ensure that the scented air could be delivered. Unlike electro-pneumatic tube

systems, this device has some points where gaps are inherently present. For example, the

selector was a particularly difficult part because the air would at the transitions between

tubes. Also, section reductions and tube lengths posed problems for the limited power of the

DC fan. Despite the issues, after some improvements the final results were satisfactory.

52

5.2 Implementation

Scent Generation

The OIKOS scents are stored as a compacted powder inside small metal cartridges (Figure

5.3) [33] that come in several sizes to suit each of the company’s products. As mentioned

before, the company produces ambient scent diffusers mainly for hotels, events and stores.

Rather than just diffusing nice smells for people to enjoy, it considers the psychological

effects of smell on people for marketing purposes and it studies the chemical interaction of

the scent and the air to also include air purification capabilities.

Figure 5.3 - OIKOS Scent Cartridges: On the left, the scented powder. In the Center, a cartridge with the compacted powder. On the right, the cartridge used in the prototype.

53

The company’s most famous product is called the Cube (Figure 5.4) [34] and it offers a

superior aroma experience. The device is very simple, it uses a USB powered dc fan to

generate and deliver the smell, diffusing the right amount of scent to make the experience

enjoyable. The scents are sold in grouped packages, like the citrus pack for example, and

cartridges can easily be replaced. It is relevant to note that the Solid Fragrance Release (SFR)

does not permeate the people’s clothes like other liquid or gas based odors do. This is also a

great advantage in the Olfactory Display since one of the usual issues is the odor particle

adhesion to the display.

Figure 5.4 - OIKOS Cube

Figure 5.5 - Erosion Process

54

The scent release mechanism depends on the airflow interaction with the cartridges. The

passing air creates an erosion effect (Figure 5.5) [35] that releases micro particles that travel

through the tubes of the Olfactory Display. The amount of particles release depends on the

surface shear stress created by the airflow, so air speed and the angle of attack are the 2 key

factors that influence scent concentration. For this reason in the developed prototype, the

cartridges are placed with a small angle between its surface and the airflow (Figure 5.6). The

cartridges used are the smallest in the OIKOS range (Figure 5.3 Right) to fit in the narrow

tubes. For each of the 3 scented tubes, there are 4 of these cartridges.

Figure 5.6 - Scented Tube: Top Left, the complete scent tube. Top Right, the scent tube without the cartridges. Bottom, Drawing view of the tube, note the section view of the tube where air flows from the right to left.

55

Scent Selection and Delivery

The Selector design was inspired by the cylinder of a revolver gun. In this case, our selector

has 4 chambers or tubes from where the air can flow: 3 are scented and one is empty to

clean the system. The rotation of the selector is controlled by a servo motor that permits a

precise alignment between the selector tubes with the inlet and outlet tubes.

Roughly explained, a servomotor (Figure 5.8 Right) [36] is a rotary actuator that allows for

precise control of angular position, velocity and acceleration [37] and it can rotate within a

180° range. It consists of a motor coupled to a sensor for position feedback. It also requires a

relatively sophisticated controller, often a dedicated module designed specifically for use

with servomotors [36].

Figure 5.7 - Olfactory Display Selector: On the top, highlighted selector in dark blue and the inlet and outlet tubes highlighted in lighter blue. Bellow, side view of the selector without the servo.

56

The Figure 5.7 presents the selector design. Air flow comes from the inlet at the left, crosses

the desired tube and then is directed to the user’s nose. Gaps were kept to a minimum

between the selector tubes and the neighboring inlets and outlets, this was fundamental to

ensure the necessary fluid dynamic efficiency. In the real prototype, the tubes don’t align as

perfectly as in this 3D representation, but the results were still satisfactory.

From the front view, one can notice that there is a considerable gap between each of the

tubes. This gap was as open as possible to avoid crosstalk between different scents. All the

tubes are placed within a 160° range so that it sits within the servomotor range. Because

servomotors are not ideal, they’re effective range is slightly below 180°, and so an extra

margin was given when placing the tubes in the selector.

After passing through the selector, the scented air is delivered to the user’s nose through a

flexible transparent plastic tube. The tube diameter had to be carefully considered: too

narrow and the air wouldn’t make it through until the end, too wide it would compromise

the user’s comfort. Also, the length of the tube had to be long enough to reach the user’s

nose and allow for him/her to move freely. Again, a longer tube means more friction

between the inner walls of the tube and the airflow.

Figure 5.8 - On the left: Front View of the selector showing the tube position within an angular range and the gaps between each tube. On the Right: A servomotor

57

The Control Unit

6.1 Design

The control unit is, as the name suggest, the sub-system designed to control the Olfactory

Display. The control had to be reliable, simple and quick to respond. The control unit is

structured into two main parts:

Computer Connected - This part (Figure 6.2) is responsible for receiving the input of

the user controlling the device and the transmitting of the appropriate signal to be

received by the device. The user-interface is an application created in the software

Processing [38] that sends the command to an Arduino [39] programmable board.

The board includes an RF transmitter, a radio based wireless system that sends the

signal to the receiving end.

Figure 6.1 - System Architecture. The Control Unit elements and their integration in the system.

58

Device Connected - This part, consists of another Arduino board (Figure 6.3) that has

the RF Receiver module connected to it to receive the signals. These signals are then

sent to the electronic devices - the DC fan and the Servo motor - that are

incorporated in the Olfactory Display.

In the development of the control unit, there were 3 essential tools that were used: The

Arduino programmable boards, the user-interface software Processing and the RF Wireless

system.

Figure 6.2 - Computer Connected unit with the transmitter board

Figure 6.3 - Device Connected unit with the board, DC motor and servomotor connected to it

59

6.1.1 Arduino

Arduino is an open-source computer hardware and software company, project and user

community that designs and produces kits for building digital devices and interactive objects

that can sense and control the physical world [39] [40]. The company was made famous due

to its starter kit that included a programmable board, the Arduino Uno (Figure 6.4), and a

few electronic sensors and actuators, along with a projects book to get beginners started in

building do-it-yourself projects. The Arduino is a highly versatile tool and because it is an

open-source initiative, it is possible to find an immense amount of fundamental information

of all sorts of projects.

For this project, it proved to be an essential tool. Each Arduino Uno board, has 14 digital

input/output pins, 6 analog inputs, a 16 MHz ceramic resonator, a USB connection, a power

jack, an ICSP header, and a reset button. To get started one just simply connects it to a

computer with a USB cable or power it with an AC-to-DC adapter [39].

Figure 6.4 - The Arduino Uno board

60

6.1.2 Processing

To design the user interface, another open-source initiative was used. Processing (Figure 6.5)

is a programming language and integrated development environment (IDE) built for

electronic arts, new media art, and visual design applications [38] [41]. Like the Arduino, the

project was also intended to teach the fundamentals of computer programming in visual

context, so that the general public could learn and develop their own projects. The program

uses the Java language with a simplified syntax and graphics programming model. With

processing, very basic interfaces and highly advanced graphics can be developed. Processing

has a very similar interface to the one of Arduino, and both programs can communicate with

one another.

6.1.3 Wireless System - RF Module

The sub-system that establishes the wireless communication is called RF Module. An RF

module consists of two small electrical components, a transmitter and a receiver, that

communicate between one another via antennas that can be incorporated to each of the

boards (Figure 6.6). Both components have a series of pins, the transmitter has 4 and the

receiver has 8, and each of them serves a specific purpose (table references).

Figure 6.5 - Processing Software Logo

61

The RF (Radio Frequency) module uses radio signals to communicate, operating around the

300-400 MHz frequency range [42]. It is commonly seen in the little remote controls that

open and close garage doors.

Figure 6.6 - RF Module with pin definitions: On the left, the Transmitter (TX) board. On the Right, the Receiver (RX) board.

Pin # Name

1 AN

2 GND

3 DA

4 VCC

RF Transmitter Pins (From left to right)

The AN or ANT pin is the antenna. In this system, a 15 cm wire was welded to the pin.

GND is the pin that must be connected to the ground.

DA or DATA is the pin that receives the data to be sent.

VCC or VIN is the pin to be connected to the positive pole of the power supply.

Description

Table 6.1 - RF Transmitter module pin definitions and purposes.

Pin # Name

1 AN

2 GND

3 GND

4 VCC

5 VCC

6 NC

7 DA

8 GND

NC corresponds to the second data pin, that is not used

DA or DATA pin that receives the messages from the Transmitter

Ground

Description

The AN or ANT is the receiving antenna pin.

Ground

The same as the Transmitter.

RF Receiver Pins (From left to right)

Table 6.2 - RF Receiver module pin definitions and purposes.

62

There are several other technologies, like Bluetooth and Wi-Fi based systems that can

transmit large amounts of data and offer a near perfect connection. But they are also

considerably more expensive (can go up to a 100€ euros against the 10€ of the system used)

but it all depends on what the application requires. RF modules are one of the cheapest

technologies available that are compatible with Arduino, and it is designed to send low data

signals which is ideal for our case where we just have to send signals that command the

receiver board to do something when a specific signal is received. Another advantage is that

the set-up process is relatively simple. It is reported that it can work up to ranges of about

100m depending on the input voltage and the amount of obstacles between the 2 terminals.

On the downside they are quite sensitive to radio noise, and may suffer interferences from

other radio systems close by.

6.2 Computer Connected Part

6.2.1 Interface

The user interface is an application developed using Processing that receives the instructions

of the person controlling the Olfactory Display. The interface uses the keyboard of a

computer, with the assigned keys switching the DC Fan On and Off and alternating between

scents (Table 6.3).

For demonstration purposes, the preliminary application developed has 2 scents options, but

more command options can easily be added.

After the commands being registered by processing they are sent to the Arduino transmitter.

The several options are programed with an if-else statement: to each command pressed, a

specific message is sent to the Transmitter Arduino, in this case in the form of numbers. The

messages are sent using a serial communication, which allows for two computers to send

and receive data to one another. In this case processing by using a library, a list of predefined

commands that can be uploaded when needed, sends data to the Transmitter Arduino

through USB.

Key 1 2 4 5

Command Scent #1 Scent #2 DC Fan On DC Fan Off

Table 6.3 - Interface keys and corresponding commands

63

6.2.2 Transmitter Board

This board, represents the bridge between the user interface and the Olfactory Display. The

board is programmed to read the processing commands, and again through an if-else

statement, sends the appropriate message to the device.

TX Pins

GND

Data

Vin

Antenna

Figure 6.7 - Transmitter board circuit layout with the Pin connections of the TX module.

64

The Figure 6.7 and the Table 6.4 explain the circuit of the transmitter unit. The messages are

sent through the digital pin 10 to the data pin of the TX module. The LED lights up while the

message is being sent, to inform the user about the start and finish of the communication.

Also, the antenna is simply a 15cm copper wire that was welded to the antenna pin of the

module.

6.3 Display Connected Part

6.3.1 Receiver Board

As the last element of the connection chain, it receives the messages from the transmitter

and issues the appropriate commands through an if-else statement.

The receiver board is slightly more complex, as it has more components connected to it. The

amount of components created power shortage issues, mainly caused by the DC fan, and

also would complicate the program script. The solution was to use an Arduino Motor Shield

(Figure 6.8), which is a second board that can be attached on top of the Arduino Uno board

that facilitates the set up process of the system.

Figure 6.8 - The Arduino Motor Shield.

Arduino Pins Connected to

5V Pos. row in Breadboard

Ground Neg. row in Breadboard

Digital 10 Data Pin of the TX module

Digital 12 Signaling Led

Table 6.4 - Transmitter Arduino pin connections

65

The Arduino Motor Shield enables us to control several components like DC motors, servos

and many others simultaneously, with dedicated plugs for each type of actuator. In addition,

it has a programing library that largely simplifies the commands allowing for a better control

and a reduced script.

RX Pins GND Data

Pin #2

Data

Pin #1

VCC VCC GND GND Antenna

Figure 6.9 - Receiver board circuit layout. The Pin connections of the RX module are described in the table above.

66

The functioning is very similar to the Transmitter Board. Data is received through pin 11 and

delivered to the servo and DC Fan. Again the LED lights up while messages are being

received. Due to the excessive amount of components, a relatively powerful power supply is

necessary.

Arduino Pins Connected to

5V Pos. row in Breadboard

Ground Neg. row in Breadboard

Digital 4 Signaling Led

Digital 11 Data Pin of the RX module

Out5 Pin Trio Servo

DC Motor Channel A Couple DC Fan

Vin-Ground Power Plug External Power Supply

67

Testing/Evaluation

After the development of the prototype and the control unit, some tests were performed.

The tests initially were aimed at the checking if the devices fulfilled the basic requirements,

and then there were some attempts to improve the performance. Firstly, the Olfactory

Display testing and results are presented, followed by the control unit.

7.1 Olfactory Display Testing

With the olfactory display, it was important to have an idea about the fluid dynamic losses.

By using a small anemometer, a device that measures the air speed with a fan, some

measurements were taken at several points of the prototype (Figure 7.1) varying the input

voltage of the DC fan from 5 to 20 Volt. These tests were performed by Prof. Mario

Covarrubias [43].

Figure 7.1 - Olfactory Display with the reference points for fluid dynamic testing: P1, at the exit of the DC fan; P2, at the exit of the selector; and P3, at the exit of a delivery tube 600 mm long.

68

In these first two plots we can immediately see a speed reduction from before and after the

selector, the main cause is the misalignment between the inlet and the selector tubes since

the gap between them is almost nonexistent. Also, there is a considerable difference

between the scented tube and the cleaning tube, which would be expected due to the

presence of the scented cartridges blocking the way.

Figure 7.2 - First 2 plots for measurements taken at the P1 and P2 points.

69

The results taken at the exit of the delivery tube for both clean and scented tube, logically a

greater airflow speed reduction is observed. In the plot bellow, the diameter of the tube is

smaller, so the speed drops even lower due to the increased wall friction on the airflow.

Note that for the initial values of input voltage, the measured value is zero. This is not true,

there is a small amount of airflow reaching the point of measurement in all cases, but the

airflow is not strong enough to counter the inertia of the anemometer fan. This also happens

in the previous set of plots, but it is less evident.

Figure 7.3 - Second couple of plots of measurements taken at P3. Above, with a tube diameter of 10 mm; Bellow, for 5 mm diameter.

70

Lastly, the final plot presents all the results compared to one another. It is clear that the

biggest drop is observed between P1 and P2, suggesting that there is room for improvement

in the interface between the inlet and the selector.

Figure 7.4 - Final plot with all the measurements taken.

71

7.2 Control Unit Testing

7.2.1 Power Tests

The high power drawn by the receiver board requires a stable power supply, so most of the

initial testing was done using a stationary plug-in power supply delivering up to 1.5 amp and

with variable voltage. But the goal is to have a unit that is both portable and wireless, so

batteries would have to be used.

To begin, some tests will be performed to measure the power drawn with the power supply.

After that, some tests are performed with common batteries, to check if they provide

enough power and how long can they last.

Power consumption with the transformer

The voltage and current drawn by the DC motor, the servo motor and RF Receiver module

was measured using a multimeter. Tests were run with 7.5 V and 9 V, the measurements

were taken while the components were in standby (Idle) or being used (Activated).

Table 7.1 - Power Consumption test results.

72

The value of the electrical power (Watt) is given by the multiplication of the current (Amp) by

the voltage (Volt). The values obtained make sense since it is known that DC motors draw

more power when the voltage increases [44].

7.2.2 Battery Life Testing

In this section, the battery life is calculated with a theoretical model and some real tests are

performed to compare. 2 common battery types were tested: One of 9V and 1200 mAh of

capacity, and second one of 4.5V and 6100 mAh.

Battery Life Calculation

In a theoretical approach, the battery longevity was predicted using the equation (7.1)11

[45]. Of course battery life depends on power consumption, which means this depends on

how the Olfactory Display is used. So for this calculation, the data is based on the simulation

performed in the Battery Life Testing section.

𝐵𝑎𝑡𝑡𝑒𝑟𝑦 𝐿𝑖𝑓𝑒 = 𝐵𝑎𝑡𝑡𝑒𝑟𝑦 𝐶𝑎𝑝𝑎𝑐𝑖𝑡𝑦

𝐿𝑜𝑎𝑑 𝐶𝑢𝑟𝑟𝑒𝑛𝑡 ,

(7.1)

𝐿𝑜𝑎𝑑 𝐶𝑢𝑟𝑟𝑒𝑛𝑡 = 𝐶𝑖𝑟𝑐𝑢𝑖𝑡 𝐿𝑜𝑎𝑑 + (𝐴𝑐𝑡𝑖𝑣𝑒 𝐿𝑜𝑎𝑑 ∗ 𝐴𝑐𝑡𝑖𝑣𝑎𝑡𝑖𝑜𝑛 𝑇𝑖𝑚𝑒) ∗ 𝐴𝑉𝐺

AVG - Numerical factor (considered to be 75%) to compensate for the reduction of

drawn power as the battery fades.

Circuit Load - Current drawn while the circuit is on standby

Active Load - Current drawn by the circuit while the device and its electrical

components are being used

Activation time - Ratio between active time and total time, which will be considered a

third of the total.

73

Using the measured load values for each battery (Table 7.2), the expected battery life in

hours is obtained (Table 7.3). As seen from the previous case (Table 7.1), the DC motor is the

component that requires the most power, and it is active for longer periods, while the servo

and the RF module are active for short instants. Nevertheless, during a simulation they have

to be used quite frequently. Logically, the battery life depends on the usage. Either way the

results for battery life make sense: the 4.5 V battery last longer due to its higher capacity and

the fact that with less voltage the DC motor draws less current.

Battery Life Testing

The goal behind the following tests was to see how the system would behave in a one hour

run. During this hour a series of commands were issued and the device’s performance

throughout the period was evaluated.

The test run is a sequence of a few DC motor commands: 10 min off followed by 20 min and

then one repetition. At the same time, every 10 minutes a random sequence of servo

commands was issued.

With the 9v battery, the first 10 minutes with the DC motor off the system worked perfectly.

When the DC Motor was switched off, the servomotor started to fail and shortly after it

Units

Battery Type 4.5 9 Volt

Battery Capacity 6100 1200 mAh

Active Load 350 500 mA

Circuit Load 60 60 mA

AVG N/A

Activation Time N/A

Circuit Current Consumption

0.75

0.333

Units

Battery Type 4.5 9 Volt

Predicted Battery Life 41.38048 6.490872 Hours

Equation Results

Table 7.2 - Current consumption values of the Olfactory Display with the 2 batteries

Table 7.3 - Battery life results using the equation (7.1)

74

stopped reacting to any commands. Apart from that, the DC motor could not be switched off

and when the simulation was restarted, the same thing happened. Clearly the issue was that

DC motor was drawing all the power from the battery, leaving nothing left to operate the

remaining components of the receiver unit. Either way, there was enough power to keep the

DC motor running for the whole hour.

With the 4.5 V battery things started off better. Clearly the DC motor was rotating a lot

slower but the system responded to all the commands, which means there was enough

power for all the components to work simultaneously. However, close to the hour mark the

dc motor control was not working so sharply.

With these battery tests it is obvious that the 4.5 V battery produced better results. The

higher capacity allows it work for longer periods and the reduced voltage avoids the DC

motor from using up all the power. One might think that the good choice would be to go for

low voltage and high capacity batteries but it is not that simple, the DC motor might be

working but it needs more power to provide a satisfactory scent delivery. In addition, using

an efficient DC motor may also improve these results.

Theoretical Vs. Practical Results

The system was not tested to check if the theoretical results were real, it would take 47

hours just to verify the 4.5 v battery test result. The thing is that battery life is not that

relevant in this case, it is how long the system can use all of its capabilities, and with these

batteries this period is not long enough.

7.2.3 Wireless Communication Tests

On this part, the effective range of communication between the RF modules was tested. This

was done by sending a sequence of 20 commands, and check how many are received. Six

test runs were performed, correlating two variables: distance and the presence of obstacles.

The quality of the communication depends of a variety of factors described previously in

section 6.1.3.

Apart from the hardware, another crucial element was the programming of the

communication software. The program algorithm largely depends on the type of information

being sent, that can go from simple high and low commands or full messages. In our case,

because we needed different commands to be recognized, the information is sent in the

form of numbers and letters that can be identified by the receiving module.

75

To send messages, a coding library specifically designed for RF modules is used. A library is a

very helpful file with a series of instructions that are designed to work with a specific

component, like our RF module, and simplify the coding. The libraries influence the speed of

communication, and since commands cannot be sent simultaneously, the responsiveness of

the communication becomes more important. In this system, two different libraries were

tested:

RCSwitch [46] - This library is commonly used on small radio remotes that control

garage doors. The programming is a little simpler than Virtualwire but on the

downside the communication is slow, it took 4 seconds to process each command.

VirtualWire [47] - Designed specifically for Arduino, the VirtualWire library showed an

increased responsiveness compared to the RCSwitch. The script is a bit longer, but it

allowed for a command to be processed under 1 second. It was immediately clear

that this was the library to be used.

The tests were communication effectiveness tests were performed indoors. A random

sequence of 20 commands was tested at 3 different distances. In one case, there was a clear

line of sight between the transmitter and receiver modules, in the other, there were

obstacles in between such as pieces of furniture or walls.

The results are the percentage of commands that were correctly received by the receiver

module. When the distance is increased, some commands do not pass and sometimes a

single command can take a bit longer to be processed. The transmitter module is powered by

a USB cable, which provides 5V to the transmitter module. If necessary, effective range can

be increased with the input voltage.

3 9 12

Yes 100% 100% 95%

No 100% 95% 95%

Distance (m)

Obstacles

Communication

Effectiveness

Table 7.4 - Communication distance test results

76

77

Conclusion and Future Research

In a field that is plenty with innovative but complex solutions, an economic, simple and

reliable Olfactory Display prototype and a wireless control unit were developed that can be

adapted for several applications.

More concretely, with the Olfactory Display there are some achievements to be pointed out.

The OIKOS SFR methodology was successfully integrated, and proved to be highly adapt to

the type of device presented. Not only it worked, the OIKOS scents had the advantage of not

adhering to the device, a problem that most developers are faced with. In addition, the scent

generation, selection and delivery functions were correctly integrated fulfilling their

respective performance requirements without compromising the design’s simplicity.

The control unit developed also met its requirements. The wireless transmission of data was

accomplished using the RF modules that although cheap, provided a reliable and quick

communication within the distances usually required for these applications. In addition, a

simple user-interface was successfully integrated within the wireless communication

process.

There is plenty of room for improvement on both parts. This Olfactory Display prototype

served to test a specific design type, to use it as a final product there it has to be adapted to

a specific application. For starters, a wearable component is needed to direct the delivery

tube to the user’s nose. Fluid dynamic losses were another key aspect with this design, this

could be further improved by improving the tube transitions. In addition, extra scented tubes

could be added to the revolver selector and the limited range of 180° degrees could be

doubled with a set of gears. This way, the range of scents would be increased without using

extra room. Another way to increase the scent range would be to add another selector after

the first one, although one would have to be careful with this approach to avoid any

additional fluid dynamic losses.

With the control unit, power was the main issue that could be improved. There was no such

problem with a plug in power supply but since the prototype is meant to be wearable, the

receiver module has to be compatible with a battery or any other portable power supply. A

more efficient DC motor would mitigate this effect. Nevertheless, a powerful portable power

supply will always be necessary.

78

79

References

[1] W. Powers, "The Science of Smell Part 1: Odor perception and physiological," Iowa

State University, University Extension, 2004.

[2] T. Nakamoto, Human Olfactory Displays and Interfaces: Odor Sensing and Presentation,

Tokyo: Tokyo Institute of Technology, Japan, 2012.

[3] Water Environment Federation, "Odor Control for Wastewater Facilities. Manual of

Practice No. 22.," Water Pollution Control Federation, Washington D.C., 1978.

[4] J. Morrison, "The sense of smell - Definitions and Structural components required,"

HumanPhysiology.academy, 2014 - 2015. [Online]. Available:

http://humanphysiology.academy/Smell/Smell.html. [Accessed 05 06 2015].

[5] L. Pavesi, Graduation Thesis - Design Olfactivo: progettare con gli odori, Milan: Scuola

del Design - Politecnico di Milano, 2012.

[6] Zara, "Green Herbs Air Freshener Sticks," [Online]. Available:

http://www.zarahome.com/tr/en/fragrance/green-herbs-c1293864p5949627.html.

[Accessed 15 June 2015].

[7] D. W. Kim, K. Nishimoto, S. Kunifuki, Y. Kawakami and H. Ando, "Development of

Aroma-Card Based Soundless Olfactory Display," in Proceeding of 16th IEEE

International Conference on Electronics, Circuits and Systems (ICECS), 2009.

[8] T. Yamada, S. Yokoyama, T. Tanikawa, K. Hirota and M. Hirose, "Wearable Olfactory

Display: Using Odor in Outdoor Environment," in Proceedings of IEEE Virtual Reality

Conference, Alexandria, Virginia, 2006.

[9] A. Kadowaki, J. Sato, Y. Bannai and K.-i. Okada, "Presentation Technique of Scent to

Avoid Olfactory Adaptation," in Proceedings of 17th International Conference on

Artificial Reality and Telexistence, 2007.

80

[10] Fisitech, "A Great Instrument To Atomize Liquid: Ultrasonic Vibrator," 4 March 2011.

[Online]. Available: http://subato.blogspot.it/2011/03/great-instrument-to-atomize-

liquid.html. [Accessed 9 June 2015].

[11] J. Yu, Y. Yanagida, S. Kawato and N. Tetsutani, "Air Cannon Design for Projection-Based

Olfactory Display," in 13th International Conference on Artificial Reality and

Telexistence, Tokyo, 2003.

[12] J. Kaye, Symbolic Olfactory Display, Cambridge, MA: Master Thesis, Massachussets

Institute of Technology, 2001.

[13] C. A. Wisneski, The Design of Personal Ambient Displays, Cambridge, MA: Master

Thesis, Massachusetts Institute of Technology, 1999.

[14] S. Ren, Ambient Displays: Information Visualization through Physical Interfaces.,

Cambridge, MA: Master Thesis, Massachusetts Institute of Technology, 2000.

[15] Wikipedia, "iSmell," 7 November 2013. [Online]. Available:

https://en.wikipedia.org/wiki/ISmell. [Accessed 23 June 2015].

[16] S. Bromley, "Games User Research," 29 January 2010. [Online]. Available:

http://www.stevebromley.com/blog/2010/01/29/no-user-testing-oops-%E2%80%93-

the-digiscent-ismell/. [Accessed 23 June 2015].

[17] J. F. Morie, The Scent Collar, Marina Del Rey, California: Institute for Creative

Technologies, USC, 2003.

[18] C. J. Watkins, "Methods and apparatus for localized delivery of scented aerosols".

United States of America Patent US6536746 B2, 25 March 2003.

[19] Y. Yanagida, H. Noma, N. Tetsutani and A. Tomono, "An Unencumbering, Localizd

Olfactory Display," in CHI 2003: New Horizons, Fort Lauderdale, Florida, 2003.

[20] S. Kawato and N. Tetsutani, "Real-time Detection of Between-the-Eyes with a Circle

Frequency Filter.," in 5th Asian Conference on Computer Vision, Melbourne, 2002.

[21] Y. Yanagida, S. Kawato, H. Noma, A. Tomono and N. Tetsutani, "Projection-Based

Olfactory Display with Nose Tracking," in IEEE Virtual Reality 2004 March 27-31,

Chicago, 2004.

81

[22] F. Nakaizumi, Y. Yanagida, H. Noma and H. Kenichi, "SpotScents: A Novel Method of

Natural Scent Delivery Using Multiple Scent Projectors," in IEEE Virtual Reality,

Alexandria, Virginia, 2006.

[23] K. Murai, T. Serizawa and Y. Yanagida, "Localized Scent Presentation to a Walking

Person by Using Scent Projectors," in IEEE International Symposium on Virtual Reality

Innovation, Singapore, 2011.

[24] T. S. Lorig, D. G. Elmes, D. H. Zald and J. V. Pardo, "A computer-controlled olfactometer

for fMRI and electrophysiological studies of olfaction.," Behavior Research Methods,

Instruments, & Computers, pp. 370 - 375, 1999.

[25] A. Mochizuki, T. Amada, S. Sawa, T. Takeda, S. Motoyashiki, K. Kohyama, M. Imura and

K. Chihara, "Fragra: A Visual-Olfactory VR Game," in SIGGRAPH , Los Angeles, California,

2004.

[26] J. P. Cater, "The Nose Have It!," Letters to the Editor, Presence, vol. 1, no. 4, pp. 493-

494, 1992.

[27] J. P. Cater, personal e-mail, 5, Inc., and US Air Force Fire Research group, Tyndall AFB,

1999.

[28] OIKOS Fragrances, "Tecnologia," OIKOS Fragrances, 2005. [Online]. Available:

http://www.oikosfragrances.com/tecnologia.htm. [Accessed 29 June 2015].

[29] T. Nakamoto, S. Otaguro, M. Kinoshita, M. Nagahama, K. Ohinishi and T. Ishida,

"Cooking Up an Interactive Olfactory Game Display," Computer Graphics and

Applications, IEEE, vol. 28, no. 1, pp. 75 - 78, 2008.

[30] Wikipedia, "Aromatherapy," 13 July 2015. [Online]. Available:

https://en.wikipedia.org/wiki/Aromatherapy. [Accessed 14 July 2015].

[31] S. A. Brewster, D. K. McGookin and C. A. Miller, "Olfoto: Designing a Smell-Based

Interaction," in CHI, Montreal, Canada, 2006.

[32] M. Dirjish, "What’s The Difference Between Brush DC And Brushless DC Motors?,"

Electronic Design, 16 February 2012. [Online]. Available:

http://electronicdesign.com/electromechanical/what-s-difference-between-brush-dc-

and-brushless-dc-motors. [Accessed 7 July 2015].

82

[33] OIKOS Fragrances, OIKOS Scent Rectangles Brochure, Milano, 2014.

[34] OIKOS Fragrances, OIKOS Cube3 Brochure, Milano, 2012.

[35] G. Steinhardt, C. Schoon, Extension Soil Conservation Education Specialists and J.

Wilcox, "Wind Erosion Concerns in Indiana," Purdue University Corporate Extension

Service, [Online]. Available: https://www.extension.purdue.edu/extmedia/AY/AY-

271.html. [Accessed 30 June 2015].

[36] Wikipedia, "Servomotor," 16 May 2015. [Online]. Available:

https://en.wikipedia.org/wiki/Servomotor. [Accessed 30 June 2015].

[37] D. Sawicz, "Hobby Servo Fundamentals," [Online]. Available:

http://www.princeton.edu/~mae412/TEXT/NTRAK2002/292-302.pdf. [Accessed 30

June 2015].

[38] B. Fry and C. Reas, "Processing Software Homepage," Processing, [Online]. Available:

https://processing.org/. [Accessed 29 06 2015].

[39] Arduino, "Arduino Home," [Online]. Available: https://www.arduino.cc/. [Accessed 29

June 2015].

[40] Wikipedia, "Arduino," 30 June 2015. [Online]. Available:

https://en.wikipedia.org/wiki/Arduino. [Accessed 1 July 2015].

[41] Wikipedia, "Processing," 31 May 2015. [Online]. Available:

https://en.wikipedia.org/wiki/Processing_(programming_language). [Accessed 1 July

2015].

[42] M. Schwartz, "Open Home Automation," 7 January 2014. [Online]. Available:

https://www.openhomeautomation.net/wireless-home-automation/. [Accessed 1 July

2015].

[43] M. Covarrubias, P. Fernandes, M. Bordegoni, U. Cugini, E. Maggioni and A. Gallace,

"Wearable Multi-Fragrance Olfactory Display for VR/AR Applications.," KAEMART

Research Group at the Politecnico Di Milano, Milan, 2016.

[44] F. Braghin, "Control And Actuating Devices Lecture Notes," Politecnico Di Milano,

Milan, 2014.

83

[45] Jenesis International, "How is Battery Life Calculated," [Online]. Available:

http://jenesisproducts.com/how-is-battery-life-calculated/. [Accessed 07 July 2015].

[46] 03 November 2013. [Online]. Available: https://github.com/sui77/rc-switch. [Accessed

06 July 2015].

[47] M. McCauley, VirtualWIre, 2013.

[48] A. Gilbert, What the nose knows: The science of scent in everyday life, New York:

Crown Publishing, 2008.

[49] "How Internet Odors Will Work," How Stuff Works.com, 05 January 2001. [Online].

Available: http://computer.howstuffworks.com/internet-odor1.htm. [Accessed 2015

June 13].

84

85

Appendix A

In this the section, the code for the interface application with the software Processing is

presented:

1. import processing.serial.*; 2. 3. Serial myPort; // Create object from Serial class 4. 5. void setup() 6. { 7. String portName = Serial.list()[0]; 8. myPort = new Serial(this, portName, 9600); 9. } 10. 11. void draw() { 12. if (keyPressed){ 13. if(key=='1'||key=='1'){ 14. fill(0); //blck 15. myPort.write (49); 16. delay (10); 17. } 18. 19. if(key=='2'||key=='2'){ 20. fill(1); //blck 21. myPort.write (50); 22. delay (10); 23. } 24. 25. if(key=='3'||key=='3'){ 26. fill(2); //blck 27. myPort.write (51); 28. delay (10); 29. } 30. 31. if(key=='4'||key=='4'){ 32. fill(3); //blck 33. myPort.write (52); 34. delay (10); 35. } 36. 37. if(key=='5'||key=='5'){ 38. fill(4); //blck 39. myPort.write (53); 40. delay (10); 41. } 42. 43. }else { 44. fill(255); //white 45. } 46. rect (25, 25, 50, 50); 47. } 48.

86

87

Appendix B

Here, the Arduino transmitter board programming script is presented:

1. //This is the transmitter, that recieves from Processing and sends to the other arduino

2. #include <VirtualWire.h> 3. 4. int myData = 0; 5. int const ledpin1 =12; 6. 7. void setup() { 8. 9. Serial.begin(9600); 10. 11. vw_set_ptt_inverted(true); 12. vw_setup(2000); 13. vw_set_tx_pin(10); // Transmitter at Digital Pin 7 14. 15. // set the digital pin for the led as output 16. pinMode(ledpin1,OUTPUT); 17. 18. }//close setup 19. 20. void loop() { 21. 22. if(Serial.available() >0){ 23. 24. myData =Serial.read(); 25. 26. if(myData == '1'){ 27. digitalWrite(ledpin1, HIGH); 28. 29. 30. char *msg2 = "1"; 31. vw_send((uint8_t *)msg2, strlen(msg2));//send byte to the receiver 32. vw_wait_tx(); // Wait until the whole message is gone 33. 34. digitalWrite(ledpin1, LOW); 35. 36. } 37. 38. if(myData == '2'){ 39. digitalWrite(ledpin1, HIGH); 40. 41. 42. char *msg2 = "2"; 43. vw_send((uint8_t *)msg2, strlen(msg2));//send byte to the receiver 44. vw_wait_tx(); // Wait until the whole message is gone 45. 46. digitalWrite(ledpin1, LOW); 47. 48. } 49. 50. if(myData == '3'){

88

51. digitalWrite(ledpin1, HIGH); 52. 53. 54. char *msg2 = "3"; 55. vw_send((uint8_t *)msg2, strlen(msg2));//send byte to the receiver 56. vw_wait_tx(); // Wait until the whole message is gone 57. 58. digitalWrite(ledpin1, LOW); 59. 60. } 61. 62. if(myData == '4'){ 63. digitalWrite(ledpin1, HIGH); 64. 65. 66. char *msg2 = "4"; 67. vw_send((uint8_t *)msg2, strlen(msg2));//send byte to the receiver 68. vw_wait_tx(); // Wait until the whole message is gone 69. 70. digitalWrite(ledpin1, LOW); 71. 72. 73. } 74. 75. if(myData == '5'){ 76. digitalWrite(ledpin1, HIGH); 77. 78. 79. char *msg2 = "5"; 80. vw_send((uint8_t *)msg2, strlen(msg2));//send byte to the receiver 81. vw_wait_tx(); // Wait until the whole message is gone 82. 83. digitalWrite(ledpin1, LOW); 84. 85. } 86. } 87. }

89

Appendix C

Lastly, the code for the Arduino Receiver board is presented:

1. // RECIEVER CODE 2. // 3. //Serial port COM6, the second one after the phone jack. 4. // 5. //Servo library doesnt work, so SerrvoTimer2 is used. With this library only ms work

with the write () 6. //command. Values of ms range from 544 to 2400 7. // 8. //DC Motor commands do not work but servo does. I did not understand the reason yet.

Maybe theres a conflict 9. //between the libraries. 10. // 11. //DC MOTOR CONTROL 12. //Using the Motor Shield, the pins will be as follow 13. // 14. //Function Channel A Channel B 15. //Direction Digital 12 Digital 13 16. //Speed (PWM) Digital 3 Digital 11 17. //Brake Digital 9 Digital 8 18. //Current Sensing Analog 0 Analog 1 19. 20. 21. #include <VirtualWire.h> 22. #include <ServoTimer2.h> 23. 24. ServoTimer2 servoMain; // Define our Servo 25. 26. int const ledpin4 =4; 27. 28. void setup() 29. { 30. 31. //SERVO SETUP 32. servoMain.attach(5); // servo on digital pin 5 33. 34. //DC MOTOR SETUP on Channel A 35. 36. pinMode(12, OUTPUT); //Initiates Motor Channel A pin 37. pinMode(9, OUTPUT); //Initiates Brake Channel A pin 38. 39. pinMode (ledpin4,OUTPUT); 40. 41. vw_set_ptt_inverted(true); 42. vw_setup(2000); // Bits per sec 43. vw_set_rx_pin(11);//Receiver at Digital Pin 11 44. 45. vw_rx_start();// Start the receiver PLL running 46. 47. 48. 49.

90

50. }//close setup 51. 52. void loop() 53. 54. { 55. uint8_t buf[VW_MAX_MESSAGE_LEN]; 56. uint8_t buflen = VW_MAX_MESSAGE_LEN; 57. 58. if (vw_get_message(buf, &buflen)) // Non-blocking 59. { 60. int i; 61. 62. digitalWrite(13, true); // Flash a light to show received good message 63. // Message with a good checksum received, dump it. 64. 65. for (i = 0; i < buflen; i++) 66. { 67. Serial.print(buf[i]); 68. 69. //Servo commands 70. if(buf[i] == '1') 71. { 72. digitalWrite(ledpin4, HIGH); 73. servoMain.write(1250);//Selects smell a 74. delay(500); 75. digitalWrite(ledpin4, LOW); 76. } 77. 78. if(buf[i] == '2') 79. { 80. digitalWrite(ledpin4, HIGH); 81. servoMain.write(2100);//Selects smell s 82. delay(500); 83. digitalWrite(ledpin4, LOW); 84. } 85. 86. //DC Fan commands 87. if(buf[i] == '3') 88. { 89. 90. 91. } 92. if(buf[i] == '4') 93. { 94. digitalWrite(ledpin4, HIGH); 95. fullPower(); 96. delay(500); 97. digitalWrite(ledpin4, LOW); 98. } 99. if(buf[i] == '5') 100. { 101. digitalWrite(ledpin4, HIGH); 102. brake(); 103. delay(500); 104. digitalWrite(ledpin4, LOW); 105. 106. } 107. }//close for loop 108. 109. digitalWrite(13, false);

91

110. 111. }//close main if 112. }//close loop 113. //you can print the data entered when debugging by adding Serial.println 114. 115. /////////////////////////////////////////////////////////////////////////////

/////////// 116. 117. void fullPower() 118. { 119. 120. digitalWrite(12, HIGH); //Establishes BackWard direction of Channel A 121. digitalWrite(9, LOW); //Disengage the Brake for Channel A 122. analogWrite(3, 255); //Spins the motor on Channel A at Full power 123. } 124. 125. void brake() 126. { 127. digitalWrite(9, HIGH); //Engages the Brake for Channel A 128. analogWrite(3, 0); 129. } 130. //End Of Code