Future Now
The IFTF Blog
RFID in 2005: Middleware session
More notes from the Commerce Department's RFID in 2005 conference. The usual caveats-- my notes, not author's own words, etc.-- apply.
Nicholas Tsougas (SRA)
RFID is an enabler, but it can't fix fundamentally bad processes. It'll cut across many business processes; data quality and security management are still issues (DoD has 700 different logistics systems); data architecture, standardization, and data flow will also be keys.
Tom Mcauliffe (Motorola)
Secure asset solutions. Focus mainly on transportation (e.g., fleet management, inermodal container visibility), distribution and logistics; serve enterprise and government customers.
Container visibility project leverages commercial benefits for security, uses standards-based open architecture, balances network sophistication vs. device sophistication.
Key market characteristics
Emerging market, varied customer needs
Wide range of deices and functions
Solutions will require event-drive, real time responsiveness
Market will rapidly scale to VLS
Tags will have to work across many platforms
"Think wireless communication system, not just middleware." The challenges here are not unlike those in cellular networks, wireless networks.
Prasad Putta (OAT Systems)
OAT wrote original software standard for AutoID Center; focus on DoD, retail, CPG, high tech.
Convergence of RFID, EPC/UID (non-line of sight, read multiple items simultaneous, read w/out manual triggering, with unique IDs for items, ability to count them).
Existing applications: WMS, MES, inventory mgmt, CRM, account management
New applications: Proof of delivery; track and trace; product authentication, counterfeit, diversion; promotions
Big gap
RFID will read things you don't want it to
won't read eveyrthing you do whant
Readers misbehave
Data captured is not contextualized and ready for application consumption
Applications may not be ready for volume and serialized data
Need consistent EPC data set across multiple sites
Denton Clark ():
RFID as an enabling technology: enables transparency of business processes; but it's only effective in a well-planned strategic implementation.
Benefits: Fast, accurate method for data collection; good for repetitive transactions relying on part/package/material numbers; manual methods can't handle the volume.
4 levels of maturity
- Identify operational requirements: design, build, test
- Establish supply chain transformation goals
- Business process enhancement application integration
- Collaborative business development
Advice for new users: Ensure equipment and service providers understand your business; don't expect RFID to solve your asset visibility issues alone. Autonomic technology implementation is a marathon, not a sprint.
Michael Ricciarideli (VeriSign)
Verisign does lots of connection of infrastructures: manages DNS, payment systems, managed security services, digital certificates.
EPCglobal network is similiar to the Web:
DNS::ONS
Web sites::EPC information services
Search engines::EPC discovery services
Security services::EPC trust services
Stephen Miles (MIT Auto-ID Center)
2 new research initiatives: Auto-ID interoperability (networking the physical world) and MIT Data Center (making sense of the data).
The more complex the network, the more you need standards.
Intelligent infrastructure consists of tags, EPC, ONL, PML, Savant.
Ted Tanner (Microsoft)
Looking at business intelligence opportunities. Think bio stuff is going to be big. Doing pilot programs, RFID lab in Redmond, doing an advisory council.
Eric Domski (SAP)
Do enterprise applications in CRM, ERP, supply chain, etc.; RFID touches on all of these. Transformational. Holistic point of view. Challenges in standards, cost and maturity, data volume, operational complexity, privacy.
"