Auto-ID (RFID) Solution Proposal
Smart Items Research Program
Richard Swan Updated July 18 2003 SAP Corporate Research Center Palo Alto [Link]@[Link]
SAP Corporate Research
Business Driven Technology Innovation
Source and Create Technology that innovates business scenarios n Leverage worldwide university and industrial research n Validate the technology and the new business processes n Impact SAPs Product Roadmap n Karlsruhe, Palo Alto, Brisbane, Sophia-Antipolis, Johannesburg, Montreal
n
CRC Palo Alto - Montreal
Primary research areas: n Advanced Customer Interfaces u Desktop and mobile multi-modal user interfaces l Customer pilot stage u Socially aware interfaces
n Smart Items Research Program u Smart Vending u Auto-ID Infrastructure u Validation of Sensor nets
Smart Items - Real World Business Visibility
Smart Items Research Program
Connecting the corporate world to digitally enabled objects that represent their goods, services and assets Digitally enabled Goods Services Assets Other environments Scalable infrastructures Effective support of existing applications New applications to support newly available information
How much cash should be in the machine?
Smart Vending
Enterprise applications and digital connection architecture for vending machines Tracking and responding to a high volume of products throughout the supply chain and beyond
Smart Items Infrastructure - Auto-ID
Will the right product be
there when my customer wants to buy?
Auto-ID as Sensor Net example
Enterprise wide Applications
Supply chain planning Asset management Recalls, promotions, integrity, etc. Distributed model of the entire supply chain Local, often specialized applications Distributed Global Coherent Model Local model is computer mirror of real world as reported by sensors
Machine readable representation of the physical world
Net
Coherent Representation Data Samples Sensors Tagged products Physical World
Local Model
Local Model
Local Model
Net
Net
Net mirror
Retail Manufacturer Distribution
Mirror from Physical to Corporate World
Auto-ID Mirror
Shelf
Software World Product data
EPC# - lipstick, date, etc. EPC serial # - Case, wt EPC# - pallet, date, etc.
Shipment
Site A Transit Site B
Large Scale Sensor Net coupled to Enterprise Apps
Enterprise wide Applications
Distributed model of world based on sensors Specialized, real time applications? Distributed Global Coherent Model Local model is computer mirror of real world as reported by sensors
Machine readable representation of the physical world
Net
Data Capture & Representation Data Samples Sensors (Actuators) Real World
Local Model
Local Model
Local Model
Net
Net
Net mirror
Different kinds of Applications
Raw data
n E.g. Light Flux n Ground moisture n Stock Price
Dedicated application
n E.g. Turn on sprinklers n Control physical processes
Analytic Application
n E.g. Plant growth n Stock sector momentum n Predict jet engine service need
Example domains
n Enterprise wide
u
Information for Business Decision
n Grain yield per acre n Future grain price n Latest time until service
Finance, Goods, services, assets, people
Business Application
n Primary products trading application n Aircraft service scheduling application
n n n n n
Industrial Commercial/Retail Environmental Security/Defense Other
Sample questions
n What information gathering and control is enabled by sensor net technology that can enhance business decisions or create new business applications? n What kinds of information and control is not already available by conventional control systems? What is missing?
u
If sensor net technology was cheap and reliable, what new kinds of data would I gather?
n Where does the potential for self-organizing, quasi-autonomous operation offered by sensors nets have special value? n Are there generalized (software) principles and needs for sensor networks that are applicable over more than a narrow domain? n What are likely to be successful domains? n What are the broad business applications enabled by sensor nets?
u
What new sensors does a CEO need to make better decisions?
n What are the barriers to broad business usage?
Lack of appropriate sensors (and actuators)? u Unreliable, power limited, communication? u Software infrastructure for data gathering? u Analytic software to utilize data? u Business applications to utilize analysis? u Competition from established technology?
u
Sensor Net - Commercial Bonanza?
what application will transform the technology into a commercial bonanza? Everyone and their aunt and uncle is interested, But its a struggle to find the business model.
n Deborah Estrin, director of UCLAs Center for Embedded Networked Sensing.
u Quoted
by Gregory T Huang MIT Technology Review