Fall_Final_Report VOICE2.pdf

download Fall_Final_Report  VOICE2.pdf

of 20

Transcript of Fall_Final_Report VOICE2.pdf

  • 7/27/2019 Fall_Final_Report VOICE2.pdf

    1/20

    Table of Contents

    I. Abstract....................................................................................................................... 1

    II. Project Final Design Plan ..................................................................................... 1

    II.1. Introduction.................................................................................................................. 1

    II.2. Design Requirements ................................................................................................... 2II.2.1. Design Objectives .............................................................................................................. 2II.2.2. Functional Requirements ................................................................................................... 5II.2.3. System Parameters............................................................................................................. 6II.2.4. Design Constraints............................................................................................................. 7

    II.3. System Design............................................................................................................... 9II.3.1. Design Approach ............................................................................................................... 9II.3.2. Critical Components ........................................................................................................ 10II.3.3. Test Procedures................................................................................................................ 12II.3.4. Anticipated Risks............................................................................................................. 13

    II.4. Financial Budget ........................................................................................................ 14II.5. Project Schedule......................................................................................................... 16

    III. Summary............................................................................................................... 17

    IV. References ............................................................................................................ 19

    Figure 1: General configuration of system..................................................................... 10

    Figure 2: General outline of project ............................................................................... 10

    Figure 3: PDA to be used ................................................................................................ 11

    Figure 4: Breakdown of Tasks........................................................................................ 16

    Figure 5: Gantt Chart for Project ................................................................................... 17

    Table A: Cost of Parts ................................................................................................. 14

    Table B: Cost of Project .............................................................................................. 15

    - i -

  • 7/27/2019 Fall_Final_Report VOICE2.pdf

    2/20

    1

    I. AbstractThe Voice Activated Remote Control project will address the need of people who

    do not like to search for the remote control or do not have the energy to walk up to the

    television or any device which makes use of a remote control. This project will aim to

    create a device which can accept audio input and will send a corresponding signal to

    another device atop the instrument wishing to be controlled to perform the required task.

    We will develop an application which will run inside a device, such as a computer or

    PDA, which will send the signal to a set-top device which we will create. At the

    conclusion of this project, we will have a set-top device which will receive Bluetooth

    signals from any device which supports Bluetooth and our software which will enable

    Bluetooth-enabled PDAs to take voice commands and transfer them to our device. By

    creating a separate set-top box, we will be able to enable the product to be compatible to

    future devices which may integrate Bluetooth.

    II.Project Final Design PlanII.1. Introduction

    Our project is a voice-activated remote control, it entails putting together a device

    that will be able to control a television set using voice commands. Instead of the

    traditional infrared remote control, we are planning on extending its transmit range by

    adding a set of Bluetooth receiver/transmitter to the system. Some type of processor,

    either that of a PDA or a DSP, will be used to analyze the voice commands given by the

    user. It will then send the command via its attached Bluetooth transmitter. At the other

    end, by the television there will be a customized Bluetooth receiver to receive the signal.

  • 7/27/2019 Fall_Final_Report VOICE2.pdf

    3/20

    2

    Finally it converts the RF signal into compatible infrared signal to be sent on a modified

    remote control.

    Although our project scope will only focus on controlling a television set, the

    project can be modified for a numbers of applications. One example is a voice activated

    garage door opener. The driver will no longer have to take his/her eye off the road to

    press a button to open his garage door. Another application would be a voice-activated

    VCR programmer, just to name a few.

    Currently, the group aims to develop a prototype using two laptops connected via

    Bluetooth. We will develop an interface for users to speak to and use a program to

    analyze the voice. The command will then transfer to a module on the TV, which then

    converts the command to infrared. After a working prototype has been successfully

    developed, we will move towards a PDA. Finally, if time permits, we will build a

    remote control using a DSP chip. The technical and financial details of this project will

    be discussed later in the report.

    II.2. Design RequirementsII.2.1.Design Objectives

    During Design VI last semester, our group did a good job of getting ahead by

    choosing and finalizing on a design project. It helps us a lot by giving us plenty of time to

    finalize our objective. We also had a lot of time to do research on the resources thats

    available to do our job. As our advisor predicted, there are a plethora of ways to

    accomplish our design goals. As we foresaw different problems and how we can

    overcome them, we have changed some of the items that we agreed upon earlier in the

    semester and our objectives are now fairly clear.

  • 7/27/2019 Fall_Final_Report VOICE2.pdf

    4/20

    3

    Our first objective is to create a voice accepting program that is user independent.

    We want to remove the concept of user training. Another objective is to have high

    fidelity. We do not want the user to say Channel 5 ten times before it actually changes

    the channel. We also want the Bluetooth sender to successfully recognize the receiver. If

    the sender finds several Bluetooth devices, it should successfully distinguish the receiver.

    The software cannot be processor intensive because if it is going to run on the users

    PDA, it cannot hog up all of the resources. We want it so that another program on the

    PDA can run parallel to ours. The set-top device that we create will have to have low

    power consumption. We want it to be small and at the same time not have the user

    change batteries ever week.

    Our final objective for this project is to create software for a PDA that accepts

    voice commands to control a device. The PDA is to have a Bluetooth module. The

    Bluetooth module will communicate with another Bluetooth module on the receiving end.

    The receiving end will be a set top device that will receive the Bluetooth signals and

    convert them to infrared.

    While working on this project, the group had to overcome many technical

    problems. Some of the technical problems that we ran into were how to actually have

    software that does not require the user to go through voice training. To overcome this

    problem we decided that we will use a user-independent speech SDK. This led to yet

    another problem, actually finding an SDK. This took a very long time. We asked

    various people as well as did extensive searches on google.com. We finally found a

    company that let us use the demo version of the SDK.

  • 7/27/2019 Fall_Final_Report VOICE2.pdf

    5/20

    4

    Before porting the project to the PDA level, we decided to do the project on two

    laptops. A major technical problem that we ran into was to actually get the Bluetooth

    devices to actually handshake with each other. We are using the SDK that came with the

    Bluetooth modules. They provided us with samples of programs that we can successfully

    run. The code of the samples, however, is very, very complicated. We are still in the

    process of breaking down the code and understanding various components of the

    program. We hope to finish understanding the SDK by the end this month.

    After speaking to our advisor, Prof. Hong Man, we have two approaches on

    overcoming this problem. We plan to either create our program at the TCP level,

    assigning static IP addresses to the Bluetooth modules. While this still requires the

    handshake phase, it makes it easier to create the program since we have a better

    understanding of programming at the TCP level. Another approach to this is to fully

    understand the SDK and use the functions given to us. This approach is more difficult to

    follow because of the complexity of the SDK.

    Finally, another major problem that the group faced is to determine the format of

    the infrared. We are still looking at documentation that we downloaded from the

    Internet. We plan to continue our research for this portion into next semester. A

    finalized solution for this problem will be devised then.

    We foresee problems with the set-top device. We will have to find a DSP with a

    TCP stack in order for us to use the Bluetooth modules at the TCP level. While

    achieving our objective, we plan to minimize any problems that we run into. We plan to

    do this by conducting research as well as asking other professors on what they think is the

    best approach to resolving the problem. If at the end of next semester, we cannot

  • 7/27/2019 Fall_Final_Report VOICE2.pdf

    6/20

    5

    complete building the set-top device, we plan to use two PDAs for our project. One of

    the PDAs will be used to send and the other to receive the Bluetooth signals. We can

    also use two laptops to demo the idea.

    II.2.2.Functional RequirementsOne of the main features of the prototype is the ability to interpret voice

    commands and translate them to commands the television or other set-top device can

    understand. This feature enables the unit to provide convenience for users because it

    avoids the hassles of searching for a remote, especially with the lights turned off. With

    the voice recognition recognizing the command no matter the speaker, no training is

    required.

    The Bluetooth module also provides benefits. With infrared, a line of sight is

    required and a close distance in necessary to transmit the command. Bluetooth, however,

    eases both these restrictions. Bluetooth, because it is like a radio technology, eliminates

    the line of sight requirement. This enables the user to place the device in a not-so-

    obvious location. In the future, if the part of our device is moved from the PDA to a

    standalone unit, it can be placed underneath a desk or couch. With a range of 10 meters

    for the current standard of Bluetooth, this device can be placed away from the television

    or set-top device and still be fully functional. Also, this standalone device can be placed

    in another room and still control a device that is up to 10 meters away, provided there is

    no outside interference.

    Another feature is that when a set-top device is created to be placed on top of all

    devices employing infrared, it will allow the set to become Bluetooth-enabled. This

    allows other devices, such as Bluetooth-enabled cell phones, to control the device. This

  • 7/27/2019 Fall_Final_Report VOICE2.pdf

    7/20

    6

    would allow functionality in the future such as muting the television or pausing the VCR

    when answering or making a phone call.

    A requirement for our voice recognition application is to work on a PDA which

    uses the PocketPC operating system. Also, because of the limited power resources for

    PDAs, our application must not consume too much power. Further, because the PDA is

    used for much more than controlling a television set, our application must work while in

    the background of the operating system. While it is not required to work while the user is

    watching a movie on the PDA or using some other processor intensive task, it should

    work while a moderately processor intensive task is running. The application also must

    be such that it will always run in the background. That is, the user need not always click

    on the program to activate it. The user should have the option, though, to deactivate it

    when it is not required. This is similar to the way wireless devices on a PDA operate;

    they are on by default but the user can deactivate them when not needed. Also, the

    Bluetooth card should operate in the same manner so as to conserve power.

    For the set-top device, the requirement is that it must be efficient in its power

    consumption. We have not tested a DSP, so we are not completely aware of its power

    requirements. That said, we hope to have the set-top device consume power at a level

    such that two AA batteries would suffice. This number will be revised when we see the

    actual power usage by the IR transmitter, DSP, and Bluetooth chip; however this is the

    power usage we are aiming for.

    II.2.3.System ParametersThe functionality of the project basically remains true to the original goal, a

    voice-activated remote control. This remote control will in essence have no line of sight

    restriction witnessed in other remote controls, this is done via Bluetooth. The fidelity of

  • 7/27/2019 Fall_Final_Report VOICE2.pdf

    8/20

    7

    this is expected to be 100% accurate. We will also have module that converts Bluetooth

    into infrared. This will allow compatibility with modern devices such as television sets.

    The conversion from Bluetooth to infrared is expected to be 100% accurate as well

    because this is a simple enough procedure.

    Another feature is the recognition of the remote control of any device that

    contains Bluetooth. This will enable the remote to automatically configure itself to any

    nearby devices. This will eliminate the need for a user to program in every device.

    The remote control will also have a voice-activated portion as well. To this

    section of the project we will attach a 90% success reliability. Most voice-activated

    software is not perfected enough for an almost perfect track record.

    II.2.4.Design ConstraintsOur design basically has two alternatives. One alternative would be to use a PDA

    as the receiving device while the other is to build a set-top device which consists of a

    Bluetooth IC, microcontroller and an infrared transmitter. There are tradeoffs for each

    alternative in terms of the amount of resources that will be utilized. For example the PDA

    approach would take a shorter amount of time to build and implement while the set-top

    device approach would prove to be more cost effective. The following paragraphs will go

    over the constraints that we anticipate.

    Since both alternatives use a PDA as a transmitter, some of the constraints which

    stems from the transmitter are shared amongst the two alternatives. The first constraint

    that we have is the interference of background noise with our voice-recognition module.

    The voice-recognition module is expected to work when the TV is on which brings noise

    to the environment. In order to overcome this problem, we had designed the system to

    recognize a keyword before accepting commands. In our case, we have used the keyword

  • 7/27/2019 Fall_Final_Report VOICE2.pdf

    9/20

    8

    TV. To give a command, one would say TV followed by a command such as

    Volume up. The other constraint would be to be having the voice-recognition module

    readily available to accept command. This contradicts with conversation of the power if

    the device was to be on at all times. One way that we have thought of would be keep the

    device in extremely low power usage mode where although the device is on, it is not

    using more energy than needed to waiting for commands to be given by the user. Another

    way to solve this problem would be to always have the device sit in a charger.

    By using the PDA alternative, two PDAs will be used in the system; one for the

    transmitter which is responsible for the voice-recognition function and the transmission

    of Bluetooth commands. The other would be used as a receiver which is responsible for

    processing the packet sent via Bluetooth and then sending the appropriate command that

    was intended to control the television via the infrared transmitter. The biggest concern for

    this design is cost since a PDA would be too expensive to be used as a remote control.

    However, we have two personal PDAs that we can do the project with and we are

    planning to use this alternative as a prototype for our design. Nevertheless, the approach

    has some of constraints that we are aware of and must be overcome. Besides cost being

    an obvious constraint, placement of the receiver next to the TV would be another. With

    the PDA, it would be difficult to mount the receiver onto the TV while the set-top device

    would solve this problem.

    With the set-top device alternative, one of the constraints that will be faced is the

    amount of time we need to design and build it. The set-top device requires the research of

    the suitable components on the market, designing a circuit board that will integrate all the

    components together and building the circuit board. We fear that we might not have

  • 7/27/2019 Fall_Final_Report VOICE2.pdf

    10/20

    9

    enough time to get all of this done within the allotted time. Our assumption is that the

    components that will be ordered will be delivered on time, if the vendors do not deliver

    the components within a reasonable time frame it will certainly add to our burden of time

    constraint. The other major constraint is the compatibility of components; we are

    assuming that the components we obtain will work with each other which it should. But

    many a time manufacturers produce products that are not conformed to standards. If we

    are unfortunate, we might not be able to get the set-top device built due to this constraint.

    One way we can overcome that is to look at several alternatives in terms of the

    components. Another way would be to purchase (if possible) all the components from

    one vendor to minimize the possibility of incompatibility.

    II.3. System DesignII.3.1.Design Approach

    The design approach is simple, yet structured. Our first goal was to simulate the

    final prototype, in essence make a prototype for the prototype. The final project will

    have a PDA taking the voice command on one end and sending it over via Bluetooth to

    another end where it will be received by a DSP and converted to infrared via

    microcontroller.

    We intend to substitute a laptop computer for the PDA. This will be simple to

    downscale. However, we also have to substitute the DSP with a laptop. In fact the laptop

    will keep the function of the DSP, microcontroller and the infrared output device. Once

    we get the system of two laptops connected via Bluetooth to correspond and change

    channels we will downscale the components.

    This downscaling of the components is the final design strategy. We will have a

    PDA taking the command, which will have a built in Bluetooth transmitter and an

  • 7/27/2019 Fall_Final_Report VOICE2.pdf

    11/20

    10

    onboard microphone. The voice-recognition software will convert the voice into a

    command that will be sent over to the receiver end via Bluetooth. This command will be

    stripped from the Bluetooth and converted to infrared using a DSP, microcontroller, and

    an infrared transmitter. The figure below illustrates the basic concept.

    Figure 1:General configuration of system

    II.3.2.Critical ComponentsThe general outline of our project is pictured below.

    Figure 2:General outline of projectThe transmitter end will reside on a PDA, which will be the most expensive

    component of our project. This cost will not be necessary to incur because we will be

    using our personal PDAs in the prototype. The PDA we will be using is the Toshiba

    e740, pictured below.

  • 7/27/2019 Fall_Final_Report VOICE2.pdf

    12/20

    11

    Figure 3:PDA to be usedBecause of the available Compact Flash slot in the PDA, we will be able to insert

    a Bluetooth card. One example of such a card we may use is an IOGEAR Bluetooth

    CompactFlash Card available at Amazon for approximately $60. The link to the device

    is: http://www.amazon.com/exec/obidos/tg/detail/-/B00008KA6P/qid=1070735459/br=1-

    13/ref=br_lf_e_13//104-5547436-0877507?v=glance&s=electronics&n=3525271.

    Another component we will require is a voice recognition SDK that we can

    incorporate into our program on the PDA. The program we have been considering is

    from Sensory, Inc. and is available as a demo version. This program is user independent

    and will run on a PDA using PocketPC, which our Toshiba model uses.

    The Bluetooth chip we will purchase has not been fully determined yet. We plan

    to purchase one that will have the Bluetooth protocol stack already built in because there

    is no reason for us to redevelop a part which will not be the most important component of

    our project.

  • 7/27/2019 Fall_Final_Report VOICE2.pdf

    13/20

    12

    For the Bluetooth to infrared converter part, we will use a DSP. One requirement

    of the DSP is that it should already have a TCP stack built in so we will not have to

    program it in ourselves. Again, it is not a very important part of the project, so we will

    look for a chip which already has it incorporated because it will also avoid errors in our

    programming. An example of a chip we will use is the Voyager TCP/IP v4/v6 by Elmic

    Systems.

    We have not investigated into the type of infrared transmitter we will require,

    however we believe a generic infrared transmitter will suffice. This is because we plan to

    implement much of the details of infrared transmission inside our DSP. The DSP will

    take care of details such as the hold time for a 1 or a 0. Because the data transmitted

    via Bluetooth will be the raw bits to send to the IR transmitter, there will not be much

    processing required by the DSP.

    II.3.3. Test ProceduresNeedless to say, thorough testing is the key to building a perfect system. In our

    testing procedures, we will test not only for functionality and correctness but also system

    performance to meet design objectives. We are doing preliminary testing every step of

    the project which will be followed by more intensive testing later on. For the preliminary

    testing step, we have tested the functionality of the Bluetooth transmission, voice-

    recognition software and the software to convert a command to its hexadecimal code.

    Once the prototype is up and running, we will be testing the system based on

    functionality and performance. Some of the things we will be testing are whether it is

    user independent, low power, high fidelity and processor utilization.

    One of the features that will be incorporated in our voice-activated remote control

    is user independency. Any user should be able use this device without having to first train

  • 7/27/2019 Fall_Final_Report VOICE2.pdf

    14/20

    13

    it. The way we plan to test this feature is to have testers of different gender, age, voice

    pitch and accent to try the device.

    Another aspect we have to test is its power usage. Obviously the less power it

    consumes the better. With the device on and initially fully charged. We are planning to

    test the device when it is being fully utilized, used regularly, used sparingly and not used

    at all. We will record the amount of time the battery will last for each of those scenarios.

    Based on the result, we will either be satisfied with the battery length or seek alternative

    to improve on this aspect.

    High Fidelity implies accuracy, a voice command should be given only once for it

    to work. To test this part, we will once again have testers of different gender, age, voice

    pitch and accent to give command to the device.

    For processor utilization, we will obtain a free utility to be run on the PDA and

    record the processor utilization history while the device is on. We will run the device

    regularly, sparingly and not at all. We will then analyze the recorded log to see how

    processor intensive our application is.

    II.3.4. Anticipated RisksWhile conducting research and designing the project, we anticipate few major

    design risks. One major risk that we might face, is to synchronize with the correct

    Bluetooth device. When we demonstrate our project, it will have to synchronize with the

    receiving Bluetooth device. If there is anyone within 10 meters with a Bluetooth cell

    phone or PDA, our device will it pick up and try to synchronize. A workaround to that is

    when the sending device finds a list of devices, it will have to check if there is a server

    running on the other device. If we use the TCP approach, we can assign static IPs to our

    device and specify it on the sending end.

  • 7/27/2019 Fall_Final_Report VOICE2.pdf

    15/20

    14

    For the first stage of the project we plan to do our project on our laptops. Then

    we will port it over to the PDAs. Finally, we plan to create the set-top device and

    eliminate the receiving PDA. Another anticipated risk is not finishing the set-top device.

    This portion is very difficult because it runs at a very low level. If this happens then we

    cannot finish our project. In this situation, we plan to use our high-level prototypes. We

    will use two laptops or two PDAs to demo our concept of a Voice Activated Remote

    Control.

    II.4. Financial BudgetThe average price of a voice activated remote control in the market is about $70. The

    estimated retail cost of our voice activated remote control is projected to be

    approximately $60. The breakdown of the costs is shown below.

    Table A: Cost of PartsName of Part Cost DescriptionMicrophone $10 The microphone will used to input the voice commands.

    Microphones are usually built into PDAs.

    Speaker $3 The speaker is necessary to output any interactivecommand to the user, also usually built into PDAs.

    Microprocessor $20 This will be used to convert the BluetoothInfrared Sender $5 This piece will be used to send infrared signals to the

    controlled device.

    Bluetooth Modules(2)

    $60 We will need two Bluetooth modules, one to send andanother to receive.

    Bluetooth Chip $8 Final chip on the reciever

    Total Cost of Parts: $166

    The cost calculated above is the approximate material cost of building the voice activated

    control. It does not include the cost of labor. The estimated amount of time invested into

    researching, building and testing the remote control is 231.

    The group is composed of four members. Each member is expected to put in an

    equal amount of time into the project. Each member put in approximately 57.75 man

  • 7/27/2019 Fall_Final_Report VOICE2.pdf

    16/20

    15

    hours into developing this product, costing about $1740. Assuming that each engineer is

    paid approximately $30 per hour, the total labor cost for the project is $6930.

    Research has taken about 96 hours or about 2880 dollars. The testing phase is

    approximately 16 hours long, costing about 480 dollars. The group is expected to spend

    61 hours on documentation.

    Table B: Cost of ProjectTask Total Man Hours CostResearch 96 $2880

    Building and Development 58 $1740

    Testing Phase 16 $480

    Documentation 61 $1830Cost of Parts -- $178

    Final Cost of Senior Design Project 231 $7108

    In a real world engineering project there are more costs associated to the final project.

    The cost of the support staff, rent, utilities, overhead, profits, traveling and

    accommodation costs are just a few examples of what needs to be taken into account.

  • 7/27/2019 Fall_Final_Report VOICE2.pdf

    17/20

    16

    II.5. Project Schedule

    Figure 4:Breakdown of Tasks

  • 7/27/2019 Fall_Final_Report VOICE2.pdf

    18/20

    17

    Figure 5:Gantt Chart for Project

    III. SummaryThe aim of this project is to create a device and software which will allow a user

    to simply speak the command they wish performed and the device the command is aimed

    at will perform it. That is, the user will speak into our program, whether in a PDA or

    another device, and the command will be transferred to a set-top device which will send

    the command to the television, for example. Essentially, our device and software will

    save people time and effort when doing a very mundane task, whether it be changing the

    channel, the volume, or when pausing or forwarding a tape. While this may seem like a

    novelty product, it can be very helpful for those who are constantly misplacing remote

    controls or are too tired after coming home from a long day of work, for example. In the

  • 7/27/2019 Fall_Final_Report VOICE2.pdf

    19/20

    18

    future, our set-top device will be able to communicate with other Bluetooth devices that

    will be around the house.

    The system is meant to tie in various technologies and eliminate the line of sight

    restriction. It will be developed using a test method of two laptops before moving on to

    the final design which will consist of a PDA, a DSP, a microcontroller, and an infrared

    transmitter device.

  • 7/27/2019 Fall_Final_Report VOICE2.pdf

    20/20

    19

    IV. Referenceshttp://www.bluetooth.com

    http://www.innotechsystems.com/voicefire.htm

    http://www.itsc.state.md.us/ITSC/delvrbles/C-5-1/C_5%20Speech%20Recognition%20in%20the%20SESA%20Call%20Center%20-%20White%20Paper.pdf

    http://www.niad.sussex.ac.uk/ezine_issue.cfm?eZineID=5

    http://www.sensoryinc.com/html/products/vetoolkit.html

    http://www.techextreme.com/perl/story/15196.html

    http://www.troygroup.com/wireless/products/wireless/docs/TROY%20TI%20DSP%20Bluetooth.pdf

    http://www.voicemethods.com/

    http://www.lirc.org

    http://focus.ti.com/catalog/docs/appsoftwarebyapplication.tsp?applicationId=2&documentTypeId=1&templateId=5681&navigationId=9426

    http://focus.ti.com/catalog/docs/thirdpartysoftwarefolder.tsp?softwareId=4070

    http://winlirc.sourceforge.net/technicaldetails.html

    Tutorials:

    Bluetoothhttp://www.tutorgig.com/searchtgig.jsp?query=bluetooth

    Infraredhttp://www.tutorgig.com/showurls.jsp?query=infrared

    Remote Control Boardhttp://www.rentron.com/remote_control/remote1.htm

    Remote Control Schematicshttp://www.rentron.com/IR-TRX.htm