Disputation
Concepts and System Structures to Support
Collaborating Everyday Items
Thomas SchochETH Zürich, 28-Jan-2005
Disputation
28-Jan-2005
Slide 2
Outline
Introduction Main contributions
– Concepts– Systems– Evaluation
Conclusions
Disputation
28-Jan-2005
Slide 3
Outline
Introduction Main contributions
– Concepts– Systems– Evaluation
Conclusions
Disputation
28-Jan-2005
Slide 4
Vision
Marc Weiser (PARC), 1991– When almost every object contains a
computer then obtaining information about the real world will be trivial
Neil Gershenfeld (MIT), 1999– Proposes a proclamation of the Bill of
Things' Rights• things have the right to have an identity,• access other objects and• detect the nature of their environment
Disputation
28-Jan-2005
Slide 5
Media Break
Picture taken from: Fleisch E, Mattern F, Österle H (2002) BetrieblicheAnwendungen mobiler Technologien: Ubiquitous Commerce. Computerwoche
Disputation
28-Jan-2005
Slide 6
Thesis
Current situation: Available middleware and ubicomp platforms do not properly support applications that make use of Collaborating Everyday Items
Goal: Facilitate the development and deployment of such applications
Thesis: The concepts and systems structures presented in this work describe and support a world of Collaborating Everyday Items in asubstantially better way than would be possible with current means
Disputation
28-Jan-2005
Slide 7
Requirements
Support of basic abilities:– Identification of smart things– Localization of smart things– Control of sensors and actuators
Support of relations:– Composition– Containedness– Location model– Neighborhood– History
Implementation and deployment– Different identification and localization technologies– Programmatic access– Real-time requirements– Data storage
Disputation
28-Jan-2005
Slide 8
Introduction Main contributions
– Concepts– Systems– Evaluation
Conclusions
Outline
Disputation
28-Jan-2005
Slide 9
Overview of Concepts
Classification of concepts– High-level concepts– Concepts for basic abilities– Concepts for smart things– Concepts for the infrastructure– Procedure of registering a smart thing– Extensions– Application logic– Lifecycle
Disputation
28-Jan-2005
Slide 10
High-level ConceptsS
mar
t Thi
ngs
Infra
stru
ctur
e
Examples for tags:• RFID transponders• Barcode labels• BT modules• GPRS modules• …
Disputation
28-Jan-2005
Slide 11
Basic Abilities - Identification
Disputation
28-Jan-2005
Slide 12
Basic Abilities - Localization
Symbolic LocationModels
Physical Positions Models
[2cm, 1cm, 0.5cm](STPP)
[47.5° N, 8.5° E,420 m](WPP)
.loc1.locA.loc1(WSYM)
Truck1.locX(STSYM)
ga
ga-1
na
na-1
ma-1
fa-1 fa ma
Smart Things (ST)
Locations (LOC)
haha-1
qa
ra
pa
sa
Disputation
28-Jan-2005
Slide 13
Concepts for Smart Things
Disputation
28-Jan-2005
Slide 14
Concepts for the Infrastructure
Smart Thing
Tag Tag Reader Tag DetectionService
controlsdetects
Representation
Thing
Location Managers
Hosting Service Home Service
attached to
executes references contacts
updates
Disputation
28-Jan-2005
Slide 15
Concepts for the Infrastructure
Smart Thing
Tag Tag Reader Tag DetectionService
O
controlsdetects
I
SL
A
LL
Representation
Thing
Location Managers
Hosting Service Home Service
attached to
I
SL
A
I
SL
AO
SA
SA
ILSA
executes references contacts
updates
Communication channels:I = Identifier S = SensorL = Location A = Actuator
Disputation
28-Jan-2005
Slide 16
Application Logic
Split between– Representation
• Executed by hosting service– Location-dependent services
• Registers itself for changes at certain location
Disputation
28-Jan-2005
Slide 17
Introduction Main contributions
– Concepts– Systems– Evaluation
Conclusions
Outline
Disputation
28-Jan-2005
Slide 18
Systems Overview
Concepts verified on three systems– Developed iteratively to
• complete the concepts• test different implementation strategies
– Proof-of-concept• Every system implements a subset of the concepts
Systems– Voxi
• Developed by T. Dübendorfer & K. Römer– Wsst– Iceo
Disputation
28-Jan-2005
Slide 19
Iceo – Infrastructure
Representation as Java objects Location-dependent services as Jini services
Object Manager
Base B
Base A
Hub
Tagged Thing
Producer
Scanner A.1
Repre-sentation
DB
Internet
detects executes
runs
has
runs
Scanner B.1
...
...
Home ServiceHub Location
Manager
Base LocationManager
Base LocationManager
Tag Detection System
Tag Detection System
HostingService
RepresentationThing/Tag
Disputation
28-Jan-2005
Slide 20
Iceo – Comm. Channels
All four modules implemented for a Bluetooth tag Identification and localization only for RFID and Barcode
Scanner
Scanning
Locator
Actuator
Sensor
Tag
Identification Module
Localization Modules
Actuator Modules
Sensor Modules
contacts
contacts
contacts
contacts
Disputation
28-Jan-2005
Slide 21
Iceo – Registered Services
Base Lookup Service
Internet
Sensor
Locator
Scanning
Actuator
Object Manager
Remote Smart Thing
Storage Provider
Event Manager
Base
Base Location
Hub Lookup Service
Hub
Hub Location
Producer Lookup Service
Producer
Producer Location
Disputation
28-Jan-2005
Slide 22
Iceo – Object Manager
Disputation
28-Jan-2005
Slide 23
Introduction Main contributions
– Concepts– Systems– Evaluation
Conclusions
Outline
Disputation
28-Jan-2005
Slide 24
Evaluation Overview
Evaluation should give answers to two questions:– Do the concepts and their implementations
actually support a developer in developing smart things applications?• Implementation of a generic supply chain
application with all three systems– How can this be efficiently implemented?
• Qualitative comparison between Jini and Web Services as underlying middleware platform
Disputation
28-Jan-2005
Slide 25
Supply Chain Application
Benefits:– Total stock visibility– Quality assurance– Additional statistical functions– Process automation
LidWaters
OpenWaters
DistributeAll SuperSeller
OnTimeDelivery
FastDelivery
Flow of goods
Disputation
28-Jan-2005
Slide 26
Flow of Goods & Information
Sto
rage
1
Che
ck-in
Sto
rage
2 Che
ck-o
ut
Che
ck-in
Sto
rage
Disputation
28-Jan-2005
Slide 27
Iceo Entities
A location-specific warehouse managementservice is registered at every location
Every object (Bottle, Box, Container, Handle)is derived from the same super class
producer.plast.com
base.ontime delivery.ch
hub.ontime delivery.ch
RFID Bar code
hub.fast delivery.ch
base.fast delivery.ch
RFID Bar code
hub.lid waters.ch
checkout. lidwaters.ch
RFID Bar code
hub.open waters.ch
checkout. openwaters.ch
RFID Bar code
hub.super seller.ch
checkin.super seller.ch
RFID Bar code
hub.distribute all.ch
checkout. distributeall.ch
storage2. distributeall.ch
storage1. distributeall.ch
checkin. distributeall.ch
RFID Bar code RFID Bar
codeRFID Bar codeRFID Bar
code Internet
Hub Tag Reader
Base Producer
Disputation
28-Jan-2005
Slide 28
Jini vs. Web Services
Voxi & Iceo based on Java/Jini Wsst based on .Net Web Services/SoapUDDI
Java/Jini .Net / SoapUDDI FactorMemory usage of runtime environment (Kbyte) 9564 22824 2.4Response time service registration (ms) 137.6 ± 18.9 421.1 ± 82.7 3.1Response time service lookup with service ID (ms) 9.0 ± 3.3 284.4 ± 21.6 31.6Response time service lookup with service name (ms) 11.2 ± 3.6 546.3 ± 129.0 48.8Response time service invocation (ms) 14.4 ± 1.7 159.1 ± 5.4 11.0Response time test application (ms) 857.5 ± 32.8 4935.6 ± 260.8 5.8
same test environment for both platforms (computer, network) comparisons of corresponding entities (lookup, invocation, …)
Disputation
28-Jan-2005
Slide 29
Introduction Main contributions
– Concepts– Systems– Evaluation
Conclusions
Outline
Disputation
28-Jan-2005
Slide 30
Related Work
Smart thing systems– Cooperating Smart Everyday Objects
• Local interaction, active tag– RAUM
• Simple location tree, no representation– Auto-ID Center/EPCGlobal
• Standards for tag protocols and identifiers– SAP Auto-ID Infrastructure
• Focus on filters, connection to SAP– Volkswagen VisuM
• Linking tag readers with central database Adjacent domains
– Cellular IP– …
Ubiquitous computing systems– Cooltown– …
Disputation
28-Jan-2005
Slide 31
Main Contributions
High-level concepts– Thing, representation, tag detection hardware, services
Concepts for basic abilities– Identifier, location model,
Concepts for smart things– Containedness, composition, neighborhood
Concepts for the infrastructure– Home service, hosting service, communication channels
Application logic– Representation, location-dependent services
Recommendations for implementations– Three Systems (Voxi, Wsst, Iceo)– Supply chain application– Quantitative evaluation
Disputation
28-Jan-2005
Slide 32
Conclusions
Concepts have been proven to be useful– Verified in three different systems– Smart supply chain application– Results of performance measurements
Iceo used as underlying system for a smart facility management framework
Subset of concepts used by Intellion AG for their RFID middleware
Top Related