Auto-ID (RFID) Solution Proposal

1 / 32
About This Presentation
Title:

Auto-ID (RFID) Solution Proposal

Description:

Kraft Foods. Procter & Gamble. Backroom at Metro Future Store ... Promotion coupons. Shelf labels. Pointers to reference material, firmware upgrades, etc. ... – PowerPoint PPT presentation

Number of Views:165
Avg rating:3.0/5.0
Slides: 33
Provided by: SAPO

less

Transcript and Presenter's Notes

Title: Auto-ID (RFID) Solution Proposal


1
Auto-ID (RFID) Solution Proposal
Business and TechnicalObjectives for Low-Cost
Smart Items Dr. Richard SwanSAP Corporate
Research Center Palo Alto Richard.Swan_at_SAP.com
2
Smart Items - Real World Business Visibility
  • Smart Items
  • Connecting the corporate world to digitally
    enabled objects that represent their goods,
    services and assets
  • Enterprise Software Challenges
  • Scalable infrastructures
  • Effective support of existing applications
  • New applications to support newly available
    information
  • Smart Items Infrastructure - Auto-ID
  • Tracking and responding to a high volume of
    products throughout the supply chain and beyond
  • Digitally enabled
  • Goods
  • Services
  • Assets
  • Where is it?
  • What is our current inventory?
  • Will the right product be there when my customer
    wants to buy?

3
Some Core Consumer Product Goals for Auto-ID
  • Timely, accurate, detailed information
  • improves the customer experience,
  • increases sales - right product in the right
    location
  • reduces costs -reduce waste and fraud
  • Reduce overhead and allowances in supply chain
  • aids regulatory compliance
  • For example
  • Reduce Out-Of-Stock 2.5 of sales
  • Reduce Theft (internal shrinkage and
    shoplifting) 1 - 4 sales
  • Reduce Diversion, Gray/Counterfeit Market -
    Up to 45 product
  • Improve Regulatory Compliance -

4
RFID applies to several product and business
process areas
http//www.sap.com/mk/get/rfid_survey
SAP survey of executive and senior-level
management from over 20 industries. N320
respondents as of 3/03.
5
Customers want to use RFID across multiple
locations and most have no visibility across
value chain
6
Public Standards
7
Auto-ID Center
  • Founded at MIT in October, 1999
  • Currently has over 100 members and new Centers in
    England, Switzerland, Australia, China and Japan
  • SAP is founding sponsor

http//www.autoidcenter.org/
Global benefits of an integrated Smart Item
network estimated to be over U.S.240 billion
annually.
Over 550 billion different items pass through the
members supply chains every year.
8
Auto-ID Center
  • Remarkably successful Industry Consortium
  • Dramatically reduced cost of deploying RFID
  • Created royalty free standards (transition to UCC
    - EAN)
  • EAN-UCC IP firm declares no patent barrier found
    for new Class 1 UHF Tags
  • Enabled new and existing suppliers
  • Tag costs 1.00 gt 0.50 gt 0.15 gt 0.05 ? gt
    ?
  • Gillette - Alien announce 500,000,000 unit order
  • RFID Reader costs - new vendors
  • 5,000 gt 1,000 gt 300 gt 100 gt ??
  • Software
  • Interchange standards and reference
    implementations
  • RFID Reader protocol for both EPC and existing
    readers likely to be adopted by most vendors
    (structure follows proposal from SAP)
  • Identity, Discovery, Configuration
  • Reading, Writing
  • Administration and Management

9
Core of Auto-ID EPC Unique Identity Required
2 21 17 24 64
2 15 13 34 64
2 26 13 23 64
8 28 24 36 96
8 32 56 192 256
8 64 56 128 256
8 128 56 64 256
  • Unambiguously globally identifies
  • Manufacturer (EPC Manager)
  • Product Class (SKU)
  • Individual Item (serial)
  • However common interchange information is not
    present Must utilize network for
  • Shipment hierarchy (Container, pallet, case,
    item)
  • Product description (class, type, shelf life,
    etc.)
  • Individual information (manufacturing date,
    batch, other properties, history )
  • Transition between GTIN and EPC extensions under
    consideration

10
Privacy Concerns
  • All consumer manufacturers and retailers are
    concerned about consumer backlash due to privacy
  • Many individuals ask about privacy
  • Auto-ID Center standard provides for permanent
    killing of tag at check-out (consumer option)
  • Not just into reversible hibernation
  • Standard has 16-bit secret key to kill tag
  • Air protocol requires full sequence of steps to
    invoke kill, that can be monitored in-store to
    detect malicious use of kill
  • However, significant value to keeping tag active
    after consumer sale in order to support returns,
    warrantees, in-home technology, etc.

11
SAP RFID Pilot Activity
12
US Pilot - Procter Gamble
PG Distribution
Retailer DC
Retail Store
BackroomFront roomSmart Shelf
Tote Building
Cross-Dock
Customer Demand
MIT-Auto-ID SavantCaptures events
MITSavant
SII provides visibility(application
agnostic) Traditional applicationsdrive business
processes
SAP SII prototypeInterprets events toprovide
business information
AdaptiveResponse
SAP Landscape
13
Opening of Metro Future Store (SAP Pilot)
  • Rheiberg Store and Essen Warehouse
  • Includes RFID tagging of most goods at case and
    pallet level
  • CDs and other goods at item level
  • Three smart shelves for item level goods
  • Provide visibility at warehouse and store level

14
Future Store Smart Shelves
Goods movements are tracked by RFID data capture
Gillette
  • Features
  • Remote control of shelf inventory
  • Control of reminaing shelf life
  • Analysis on Goods movements from and to shelf

Kraft Foods
Procter Gamble
15
Backroom at Metro Future Store
16
RFID/Auto-ID Infrastructure
17
Core Issues for Auto-ID and Enterprise Software
  • Key mismatch 1
  • Tag readers provide notification events when tags
    are sighted
  • Locally thousands/sec, nationally millions or
    billions of events
  • Enterprise Software do not want low level
    notification events!
  • Business decision makers and Enterprise software
    want clean, stable information about the location
    and status of every tagged object and the
    relationship of those objects to other objects
    and to business documents and processes.
  • Key mismatch 2
  • Tagged objects originated from many different
    manufacturers who expect their own enterprises to
    benefit from the tag-reader beeps from throughout
    the supply chain.
  • Most Enterprise Software applications are geared
    to support a single enterprise and not oriented
    to share information with other enterprises
  • Key mismatch 3
  • Most Enterprise applications are not designed to
    deal with huge numbers of serialized items.
  • Most Enterprise Systems will fail if given 106
    more data.
  • Most Enterprise applications do not need serial
    number specific data but certain core uses of
    RFID tags, such as item history tracking, proof
    of ownership and product recall depend on serial
    number level information.

18
Fully Exploit Opportunity of Auto-ID
  • Cross Enterprise Information
  • The big news with Auto-ID is the opportunity to
    share detailed information throughout a supply
    chain in a retailer driven way
  • gt Design system from the core to share
    information both within an enterprise and between
    enterprises to gain maximum return on auto-id
    investment
  • Recognize Variety of Landscapes
  • Manufacturers sometimes have different Enterprise
    landscapes for each brand or category
  • Retailers also have a wide variety of landscapes
    and business processes which vary by category and
    supplier relationship
  • gt Must provide information in the easiest way
    to absorb and be Application Agnostic
  • Support Business Continuation at each level
  • Retail sites often lack good telecommunications
    and IT support
  • gt Enable local operation through network
    outages and unsupervised recovery from power and
    other failures
  • Get full information value of Auto-ID
  • Individual items can be tracked at serial number
    level
  • gt Provide distributed object system to
    represent each item

19
New Capabilities from Auto-ID - Item Level
  • What information is stored at the item level?
  • Items are no longer generic.
  • Create, store and update information at the item
    level and make it accessible across the supply
    chain
  • Handle promotions at item level
  • EPC discount0.50 promotion code567
  • Support customer loyalty, warrantee and life
    cycle of item
  • EPC sold-retail location987 5/9/02 loyalty
    card23458967 price 35050
  • Support duty, taxation and other regulatory
    issues
  • EPC duty_paid date, ref,R CA_tax ref,
    inspected XXX
  • Manage returns with convenience and minimum fraud
  • EPC sold-retail location967 5/8/02 returned
    6/3/03 refund 27.50
  • Reduce shrinkage through detail tracking and
    legal prove of ownership
  • EPC missing - last seen 2/7/03 truck-driver
    J.Smith
  • Targeted recalls at item level
  • EPC anti-skid breaks, batch345 recall 6/3/03 -
    unreliable electronics

20
Where and Auto-ID Infrastructure Fits
21
Core Auto-ID Node
  • PML/XML
  • Master Data
  • Item specific data
  • Distributed synchronization
  • Plug-Play
  • Central configuration and monitoring
  • Filtering at reader level
  • Full support for writable tags
  • Cached on demand
  • Master data
  • Item level information
  • Other related info

Wired and Wireless Network
22
Basic Item level information flow
23
Cross Industry Information Flow
24
Does it Scale?
25
Brute Force Approach
  • Assumptions
  • 10,000 Stores
  • 10 readers each
  • 100 items/sec
  • 100 Bytes per lookup
  • All taken to a Central Repository
  • gt 10,000,000 lookups per secondgt 8 Gbits per
    secondgt 40 Terabytes /day (11 hours)

Wrong Approach!
26
Approach for High Volume Individual Item Tracking
  • Caching at each level dramatically reduces
  • data bandwidth to repository
  • Enables item level tracking and recording

Core Version
Cache
Cache
Cache
Node
Node
Node
Node
Node
Node
Node
Retail
Manufacturer
Distribution
27
Data Volume ( 1 billions items/year)
  • Assumptions
  • Items per year 1,000,000,000
  • gt 4,000,000 items per day (255 working
    days)
  • gt 360,000 /hr (11 hrs/day)
  • gt 100 items/sec sold
  • Levels of distribution 3 ( of sites touched by
    each item)
  • 100 Bytes per update
  • Aggregate data bandwidth (at top level) with good
    caching
  • 100 items 3 levels 3 synchronization 100
    Bytes
  • gt 90 Kbytes/sec
  • 0.72 Mbits/sec ( Less than 1 T1)
  • Store 1 KByte per Item
  • gt 4 G Bytes/day of shipments (lt 20 day )

28
Concluding Thoughts
29
World Changes One Step at a Time
  • Usage mode and adoption of Auto-ID will vary
    widely
  • Specific value to product categories and
    individual corporations
  • Market power of beneficiaries
  • Meeting regulatory mandates
  • Price curves of technology
  • Basic Building Block for long term Architecture
  • Core local, real time Auto-ID node
  • Also supports traditional RFID, writable tags
  • Pathway for flexible adoption
  • Single Enterprise
  • Extended Enterprise -
  • Cross Enterprise
  • Added dimension of granularity and business need
  • Pallet
  • Case
  • Generic Item
  • Individual item with full history

30
Billing and Reconciliation
  • As Auto-ID becomes accepted as reliable evidence
    of transfer of goods
  • At case and Pallet level, large retailers will
    pay on goods receipt, rather than through billing
  • At item level, consignment sale, or pay-on-sale,
    will become much easier.
  • Retailers may merely lease out shelf space and
    take no other financial risk

31
Summary
  • Auto-ID can have a BIG impact on supply chains,
    and operating practices in warehouses and
    particularly at retail.
  • Auto-ID using inexpensive RFID technology is
    backed by very powerful market forces (Wal-Mart
    alone is 244B net sales)
  • The Auto-ID driven software and hardware
    standards will eventually drive out older, more
    expensive RFID products and standards(20
    billion units forecast in 5 years)
  • The challenge is to make effective use of the new
    data
  • The SII (Smart Items Infrastructure) prototype
    provides the basis for large scale Auto-ID
    deployment and is the core of SAP product
    developments.
  • An expansive long term vision is need to get the
    full business value out of Auto-ID

32
Thank You
Write a Comment
User Comments (0)