ESD.290
Special Topics in Supply Chain
Management
Brian Subirana & Sanjay Sarma, MIT
History
z
z
z
z
z
1998: DISC founded
1999: Auto-ID Center founded Auto-ID Field Trial
started 2000
2001: First standards presented
2002: Gillette orders 500,000,000 tags from Alien
2003: Wal-Mart, DoD Mandates
EPCglobal launched, Center retired
z
2004: More mandates
Outline
z
z
RFID and the Auto-ID Center
An in-depth look at some issues
Outline, Part I
z RFID
z
z
z
z
z
and the Auto-ID Center
What and why of RFID
The cost issue
Manufacturing low-cost RFID
Handling the data
Current status
z An
in-depth look at some issues
Outline, Part I
z RFID
z
z
z
z
z
and the Auto-ID Center
What and why of RFID
The cost issue
Manufacturing low-cost RFID
Handling the data
Current status
z An
in-depth look at some issues
Magnitude of Challenges
z
Inventory Management:
z
Inventory uncertainty:
z
z
z
Stock-outs:
z
z
z
Average 9% out of stock in retailers world-wide
Lost sales due to stock-outs: 4%
Overstock: Huge channel inventories
z
z
z
z
65 % of 370,000 records inaccurate (HBS study of one major retailer)
Transportation uncertainty: Perfect delivery is dismal
CPG average 11 weeks inventory
Retailers average 7 weeks inventory
Locked up capital, industry-wide
Brand Management:
z
Counterfeit:
z
z
$500B pharmaceuticals business, $50B counterfeit
Diversion:
z
Market size difficult to estimate
The problems are everywhere
OOS Averages by Category
Overall OOS Extent (Averages)
8.3
World Avg (18 categories )
Wor ldw ide
8.3
Salty Snacks
Other Regions
8.2
Toilet Tissue
8.6
Europe
7.9
USA
0.0
2.0
4.0
6.0
8.0
5.3
6.6
Fem Hygiene
6.8
Diapers
7.0
7.7
Laundry
10.0
9.8
Hair Care
0.0
Percent OOS
2.0
4.0
6.0
8.0
10.0
12.0
Percent
Across geographies
OOS by Day of Week
(Average of 13 studies)
Other Cause
4%
10.9
Mon
10.0
Tues
Retail HQ or
Manufacturer
14%
9.1
Store Ordering
34%
8.7
Fri
Sat
Store Forecasting
13%
Distribution Center
10%
9.8
Wed
Thur
Across product lines
7.3
10.9
Sun
0.0
2.0
4.0
6.0
8.0
Percent
Across time
10.0
12.0
Store Shelving
25%
Across the supply chain
The Vicious Cycle
Upstream In-store
causes causes
28%
59%
Demand
variations
13%
14%
Upstream
problems
50%
Problems at
manufacturer’s DC
Store
ordering
35%
Problems
at retailer’s DC
Manufacturer
Manufacturer’s DC
Retailer’s DC
Shelf
Replenishment
Retailer
11 weeks inventory
6 weeks inventory
Consider stock-outs
8%
World
Wide
RFID System
RFID
Low cost rfid
Silicon: 4c/mm2
handling cost
die size/cost, cents
20
15
10
5
time
Why is RFID expensive today?
reduce functionality
(Networking & software)
greater functionality
reduce chip size
(handle small chips)
increased chip size
Noise
Cheap protocol
100
90
80
70
60
50
40
30
20
10
0
905
Gaussian Fit 0.8
912 100
2 Meters, 4 KHz,
SQW, 1 dBm
910
915
Frequency (MHz)
920
the hypothesis or bet
z
Place unique number on tag
z
z
z
z
Electronic Product Code, EPC
64 bit, 96 bit, and upwards
01. 203D2A. 916E8B. 8719BAE03C
Header 8 bits
Serial Number 40 bits
Product 24 bits
Manufacturer 24 bits
Develop manufacturing technology
for small chips and tags
Move data on the network
z
z
Network service for resolving EPC
Network architecture for gathering
and routing data
context-aware
context-aware
router
context-aware
router
context-aware
router
router
context-aware
context-aware
context-aware
router router
router
context-aware
router
context-aware
context-aware context-aware
router
router
router
sensor
sensor
sensor sensor
Outline, Part I
z RFID
z
z
z
z
z
and the Auto-ID Center
What and why of RFID
The cost issue
Manufacturing low-cost RFID
Handling the data
Current status
z An
in-depth look at some issues
Low cost RFID
IC IC
Design
Design
$X Million
IC
Manufacture
Antenna
Manufacture
Antenna/IC
Assembly
Conversion
to Package
20¢
5¢
5¢
20¢
1-2¢
1¢
1¢
1¢
Billions Number of tags
End
users
Challenges of IC minimalism
z 0.25
z
mm2: does it make life tougher?
……..
z
z
z
z
z
z
z
z
z
z
z
z
z
z
z
z
z
z
z
Street width will dominate
Still have to test the IC’s (?)
Die handling costs are high
Die-attach/wire-bonding techniques do not scale
Street width will dominate
Still have to test the IC’s (?)
Die handling costs are high
Street width will dominate
Still have to test the IC’s (?)
Die handling costs are high
Street width will dominate
Still have to test the IC’s (?)
Die handling costs are high
Street width will dominate
Still have to test the IC’s (?)
Die handling costs are high
Street width will dominate
Still have to test the IC’s (?)
Die handling costs are high
low cost rfid challenges
IC
Manufacture
Antenna
Manufacture
Antenna/IC
Assembly
Conversion
to Package
20¢
5¢
5¢
20¢
1-2¢
1¢
1¢
1¢
Testing
z Economics
z
$500 - $1000 per wafer
z But
z
z
z
today:
minimal functionality means
High reliability
Don’t test on wafer
Test wirelessly at conversion
Slicing and Dicing
z
Standard
saw-dicing wasteful
z Instead, use separation
by thinning
C. Landesberger, S. Scherbaum, G. Schwinn, H. Spöhrle: “New Process Scheme for Wafer Thinning and Stress-free Separation of Ultra Thin
IC’s,” Proceedings of Microsystems Technologies 2001, Mesago, Stuttgart, pp. 431-436, 2001.
Low cost RFID challenges
IC
Manufacture
Antenna
Manufacture
Antenna/IC
Assembly
Conversion
to Package
20¢
5¢
5¢
20¢
1-2¢
1¢
1¢
1¢
Antenna
z
z
z
Screen printing
Etching
Forming
Low cost RFID challenges
IC
Manufacture
Antenna
Manufacture
Antenna/IC
Assembly
Conversion
to Package
20¢
5¢
5¢
20¢
1-2¢
1¢
1¢
1¢
Assembly
z
z
z
Fluidic Self Assembly
Vibratory Assembly
Pick and place
Vibratory Assembly
Chip
Assembly
Chip
Design
Wafer
Treatment
Silicon
Manufacturing
Antenna
Manufacturing
Label / Tag
Manufacturing
Inlet
Assembly
Label
Converting
End User
Low cost RFID challenges
IC
Manufacture
Antenna
Manufacture
Antenna/IC
Assembly
Conversion
to Package
20¢
5¢
5¢
20¢
1-2¢
1¢
1¢
1¢
Conversion
z
z
z
Paper/package/label industry expertise
Scales well with mass production
Capital equipment expenditure
software
hardware
paper
Outline, Part I
z RFID
z
z
z
z
z
and the Auto-ID Center
What and why of RFID
The cost issue
Manufacturing low-cost RFID
Handling the data
Current status
z An
in-depth look at some issues
Architecture: Local
Middleware
Reader
Local network
Data
Processing
Reader
01.203D2A.916E8B.8719BAE03C
Tag
Local database
Architecture: Global
ONS
Internet
PML Server
<EPC= 01.203D2A.916E8B.8719BAE03C >
<TEMPERATURE=15 DEG C>
</EPC>
</TIME>
</PML>
Local system
Local network
Reader
XQL PML
18.72.100.100
Middleware
Reader
18.72.100.100
<PML>
<TIME=2000.4.28:10:05.05HRS>
Data
Processing
01.203D2A.916E8B.8719BAE03C
Local database
<PML>
…
????
…
...
<MaximumTempearture>
40 DEG C
</MaximumTempearture>
<MinimumTemperature>
5 DEG C
</MinimumTempearture>
Quality
control specialist
</PML>
01.203D2A.916E8B.8719BAE03C
Inference
Temperature OK
ONS
Internet
PML Server
18.72.100.100
<PML>
<TIME=2000.4.28:10:05.05HRS>
<EPC= 01.203D2A.916E8B.8719BAE03C >
<TEMPERATURE=15
</EPC>
</TIME>
</PML>
Local system
Local network
Reader
Data
Processing
<PML>
…
…
...
<MaximumTempearture>
01.203D2A.916E8B.8719BAE03C
????
40 DEG C
</MaximumTempearture>
<MinimumTemperature>
01.203D2A.916E8B.8719BAE03C
23AB.36C2.AB21.6733
Tag
XQL PML
18.72.100.100
Middleware
Reader
DEG C>
Local database
5 DEG C
</MinimumTempearture>
Quality
control specialist
</PML>
Three Layers of an EPC
Architecture
Enterprise
Edge
Existing Systems / EAI
ONS
TradingPart
Partners
Trading
ners
Trading
Partners
EPC
Enterprise
Business IDs
+
EPC Edge
object
time
loc
Edge 3
Edge 2
Edge 1
•EPC Enterprise data store
•Data migration from Edge
•Standard API
•Enterprise systems
•Trading partners
•RF Abstraction layer
•Device mgmt
•Event mgmt
•EPC Edge data store
•Standard API
•Data migration to Enterprise
Software
Hardware
Readers
Tags
•Physical data capture
Outline, Part I
z RFID
z
z
z
z
z
and the Auto-ID Center
What and why of RFID
The cost issue
Manufacturing low-cost RFID
Handling the data
Current status
z An
in-depth look at some issues
Field Trial
P & G FACTORY
CAPE GIRADEAU , MO
COCA COLA BOTTLER
CLEVELAND, TN
JOHNSON & JOHNSON
OLIVE BRANCH, TN
KRAFT FOODS
FORTH WORTH, TX
UNILIVER DIST. CENTER
BALTIMORE, MD
PILOT TEST FACILITY
BENTONV ILLE, AR
WAL-MART STORE
BROKEN ARROW, OK
WAL M ART STORE
CLEVELAND, TN
Warehouse
Pilot facility is being used
as a mini warehouse
Retail Floor
GILLETTE DIST. CENTER
CHICAGO IL
P & G DC
IOWA CITY, IO
WAL-M ART DEPOT
BENTONV ILLE, AR.
SAM'S DEPOT
KANSAS CITY, M O
SAM'S STORE
TULSA
Staging Area
Retail Floor
The Commercialization of EPC
z
z
z
Landmark Event: EPCglobal is formed
Many companies have significant tests and pilots underway
Mandates:
z DoD
Marks & Spencer
z Tesco
z Wal-Mart
z Metro Group
z Target
z Albertsons
z Best Buy
Other major retailers are continuing to announce their strategies
z
z
RFID Status
z
3 protocols
z
z
z
z
Class 0 UHF
Class I UHF
Class I HF
Tens of manufacturers
z
z
Tags: Alien, Matrics, Philips, ST Micro, Rafsec, ….
Readers: Alien, Matrics, AWID, ThingMagic, Tyco, Symbol,
Samsys,…
z
New versions being designed
z
z
Gen 2 taking off
Intermec patent still issue
Key philosophy #1:
interoperability
Sensors
Open tag
Agile reader
Software
Internet
Key philosophy #2: Layers
Class IV tags:
Active tags with
broad-band peer-to-peer communication
Class III tags:
semi-passive RFID tags
Class II tags:
passive tags with additional
functionality
Class 0/Class I:
read-only passive tags
Downward failsafe
Upward compatibility
Class V tags
Readers. Can power other Class I, II and III tags;
Communicate with Classes IV and V.
Vendors
z
z
z
z
z
Chips: Alien, Matrics, Philips, ST Micro
Readers: Alien, Matrics, Philips, Tagsys, Samsys,
ThingMagic, Tyco, Symbol, Markem, AWID
Software: Sun, Oat Systems, Manhattan, Globe
Ranger, Conecterra, SAP, Tibco, Verisign, Vizional,
..
Systems: Accenture, PWC/IBM, GEA, …
End-Users: Gillette, Wal*Mart, P&G, TESCO, Metro, Target, Wegmans, ….
Research Issues
z
z
z
z
z
z
z
Tag anti-collision
Reader anti-collision
Security and privacy
Advanced sensor networks
Data routing and handling
IC Design
IC manufacturing
z
z
z
z
z
z
z
Silicon processing
Chip assembly
Polymers
Controls/automation
Manufacturing systems
System Synthesis
Supply chain issues
Outline, Part I
z RFID
z
z
z
z
z
and the Auto-ID Center
What and why of RFID
The cost issue
Manufacturing low-cost RFID
Handling the data
Current status
z An
in-depth look at some issues
Outline, Part II
z
z
RFID and the Auto-ID Center
An in-depth look at some issues
z
z
z
z
z
A peek at the protocol
Security and Privacy Issues
Software
Vibration analysis
Silicon manufacturing
Components
z
z
z
Signaling
Anti-collision
Functions
reader to tag: modulation
Reader to Tag Modulation
UHF
Envelope
Low Time
10%
45%
100%
45%
On Off Keying (OOK), Min 90% Modulation Depth
Modulation: reader to tag
16
us
62.5 Khz
Clk Low interval for
data = 0
Clk Low interva l for
data = 1
tag to reader
1 Bit Cell
0
1
••Bit
BitCell
CellTime:
Time:~8
~8µs
µsTag
Tag
to
toReader
Reader(128
(128kbs)
kbs)
••22Transitions
Transitions==00
••44Transitions
Transitions==11
••Always
AlwaysTransitions
Transitions
Within
WithinaaBit
Bit
Anti-collision
z A
Reader Talks First (RTF) System
z Commands Issued from Reader
z Tags Reply at a Later Time While Reader
Listens
z Transactions are Self-Contained Operations
(Minimal Persistent State Information
Required)
Contention Detection
Anti-Collision Algorithm Relies on Detecting
Contention (When More than One Tag is
Responding to a Reader Command).
Contention- Two
Tags, Same Clock
Rate, 1-Bit Difference
100%
0
1
0
1
1
0
0
0
1
Anti-Collision
Query 1
1
0
00
001
000
10
01
010
100
011
11
101
110
0111
01100
011000
000
01110
01101
011001
001
011010
010
011011
011100
011101
011
100
101
01111
011110
110
00001000 (Ping)
00000000
00000000 (0)
= 0
[CMD] =
[PTR] =
[LEN] =
[VALUE]
00001000 (Ping)
00000000
00000011 (3)
= 011
111
Query 2
0110
[CMD] =
[PTR] =
[LEN] =
[VALUE]
011111
111
Functions
z
z
z
z
z
z
z
z
z
Write address
Lock address
Preload address mask
Read ID (anti-collision)
Read payload
Write payload
Sleep
Wake
Destroy
Outline, Part II
z
z
RFID and the Auto-ID Center
An in-depth look at some issues
z
z
z
z
z
A peek at the protocol
Security and Privacy Issues
Software
Vibration analysis
Silicon manufacturing
Does protocol compromise
privacy?
Not necessarily. Your choice.
z
You can destroy the tag and opt out
z
or
z
You can keep tag for later use
(physics is your friend)
z
Mass hijack of tags
z Could
happen in destroy or re-programming
z Physics
z
z
z
z
our friend
Bandwidth limited: 100’s of tags a second anti­
collision
Destroy must be individually addressed
So it takes time to kill
Surveillance
Issues
z
Tags are light-weight
z
Anyone can read the tags (promiscuity)
The same number shows up all the time
Channel is open and shared
z
z
01. 203D2A. 916E8B. 8719BAE03C
Header 8 bits
Serial Number 40 bits
Product 24 bits
Manufacturer 24 bits
Problem: unique and
promiscuous
Kill Serial number?
z
z
Product still readable
Person can be tracked by constellation
Personalize the number?
z
z
Repeated reads yield same number
You could still be tracked by constellation
Check out EPCglobal
Public policy
www.epcglobalinc.org/public_policy/
public_policy_guidelines.html
Outline, Part II
z
z
RFID and the Auto-ID Center
An in-depth look at some issues
z
z
z
z
z
A peek at the protocol
Security and Privacy Issues
Software
Vibration analysis
Silicon manufacturing
Outline, Part II
z
z
RFID and the Auto-ID Center
An in-depth look at some issues
z
z
z
z
z
A peek at the protocol
Security and Privacy Issues
Software
Vibration analysis
Silicon manufacturing
Outline
z
z
Introduction
Kinematics
z
z
z
z
z
In-plane surface motion
Out-of-plane motion
Adhesion and fluid effects
Experiments
Conclusions
Introduction
Motivation z
Need methods for handling of small
microscopic parts in tag production
processes
z
A traditional technique to solve scaling
problems is parallelization
Topics
Basic kinematics of vibratory part
transport
Effects of fluid, and
surface/adhesion forces
Outline
z
z
Introduction
Kinematics
z
z
z
z
In-plane surface motion
Out-of-plane motion
Adhesion and fluid effects
Conclusions
Vibration Kinematics
z
z
Want relative motion
Approaches:
z
Surface micro-features
z
z
z
z
Electrical/magnetic fields
Moving Fluid medium flowing over surface
z
z
MEMS cilia, rollers, etc.
Gas flow nozzle arrays
Fluidic Self Assembly (FSA Alien Tech’s)
Vibrating surfaces
z
z
Time asymmetric in-plane vibrations
Out-of-plane vibrations (such as in Bowl Feeders)
Vibrating Surfaces
In-plane vibrations of a surface if time-
asymmetric
z
z
z
Part moves if surface acceleration is more than
friction
Forward-backward motion?
Much literature
Example: Stick Slip
v
Surface
accelerating
faster than
friction can
sustain
x
Gravity
friction cone
Out-of-plane vibrations
z
z
z
z
Vertical vibration: extra degree of
freedom
Create hop, move platen back
Small-scale forces
Increased accelerations may be
needed to compensate
Hopping example
Possible return path
Part in collision
with surface
Gravity friction
cone
Part in flight
Surface accelerating faster
than gravity downwards
Outline
z
z
Introduction
Kinematics
z
z
z
z
In-plane surface motion
Out-of-plane motion
Adhesion and fluid effects
Conclusions
Adhesion
z
z
Adhesion effects become important at small
scales
Van der Waals forces
z
z
z
z
z
Due to static and quantum mechanically induced dipoles
Strong role in inter-molecular and surface phenomena
Become important at < 100nm surface separation
Adhesion surface energies of ~100mJ/m2
Clean atomically smooth surfaces in contact may have
adhesion pressures of the order of thousands of
atmospheres
Adhesion effect on vibratory
transport
z
Larger accelerations are needed z
Too contaminated or too clean are bad
z
Can be alleviated by: Roughening or Surfactants
z
Strong secondary excitation can “levitate” the parts
z
If done for fractions of the cycle, “creeping” transport can
be achieved, where the part moves when vibrations are
on and sticks when they are off
Example: hopping with
adhesion
Sticky
collision
adhesion
friction cone
Surface
acceleration
Flight after
unstick
Fluid dynamic scaling
z
z
Sticking due to
fluids
Alleviation
z
z
z
use surfaces that
are “leaky”
Make chip
surface bumpy
Run in vacuum
velocity loss due to film
damping
Frequency
effect on
damping
Si disk
400 µm dia.
200 µm thick
Starting
gap effect
on
damping
Outline
z
z
Introduction
Kinematics
z
z
z
z
In-plane surface motion
Out-of-plane motion
Adhesion and fluid effects
Conclusions
Conclusions
z
Basic physics
z
Ongoing work:
z
Measurements
z
Effects of geometries
z
Test methods
z
Design:
z
Chip delivery methods and roll to roll packaging
Outline, Part II
z
z
RFID and the Auto-ID Center
An in-depth look at some issues
z
z
z
z
z
A peek at the protocol
Security and Privacy Issues
Software
Vibration analysis
Silicon manufacturing
RFID manufacturing simulation
Gita Swamy
Sanjay Sarma
Outline
Components of an RFID tag
z Understanding the Experiment
z Manufacture
z
zSemiconductor
zTag
Results
z Worldwide Fab Capacity
z
Results
z
Total RFID Cost
zIC
+ Traditional Assembly: 4.351¢
zIC + Flip Chip Assembly: 3.311¢
IC Cost: 1.151¢
z Antenna Cost: 11¢
z Assembly: 2.25c ~ 1.15 1¢
z
z
At 10’s of billion tags a year
RFID Tag Components
z
z
z
Antenna
Mixed-Mode IC
Packaging
Inputs to model
Process Steps
Equipment Benchmark
z Throughput
z Raw
material & Utilities
z Labor
z Yield
Overhead
Maintenance
Depreciation
Semiconductor Processing
Process modeling
z
R. Leachman, J. Plummer & N. Sato-Misawa, “Understanding
fab economics” Competitive Semiconductor Manufacturing,
University of California, Berkeley, 1999.
z
Leachman & Hodges, “Berkeley Semiconductor
Manufacturing” IEEE Transactions on Semiconductor
Manufacturing, May 1996.
z
J. Bloomsburg, “RFID Tag Manufacturing” MIT UROP, 2002.
z
R. Wright, “Cost Resource Model Detail” Economic Model
Workshop, International Sematech, 2001.
semiconductor process
modeling
z Sematech
developed benchmark
z 250_A1_82 Process Benchmark
z
z
z
z
z
0.25 micron
282 Step
19 Mask
3 Metal
2 Poly
Driving Variables
z
z
z
z
Mask, Metal and Poly
Layers
Wafer starts: 300,000 per
year
100,000 dies/wafer
Wafer size: 200mm
Assembly Process
z
Assembly Process Steps
z
z
z
z
Thinning
Dicing
Assembly
z Traditional
z Flip Chip
z Fluidic Self Assembly
z Vibratory Assembly
Tag Test
Flip chip
Assembly Process
z
Types of Runs
z Unit
Machine
z Line Maximized
Results
z
Total RFID Cost
zIC
+ Traditional Assembly: 4.351¢
zIC + Flip Chip Assembly: 3.311¢
IC Cost: 1.151¢
z Antenna Cost: 11¢
z Assembly: 2.25c ~ 1.15 1¢
z
z
At 10’s of billion tags a year
Results: Die Cost
Results: Assembly Cost
Breakdown
Results: Cumulative Assembly
Cost
Assembly Costs
1.00000000
0.10000000
Traditional
Assembly
0.032175597
0.036584412
0.01000000
0.02922279
0.03025432
0.00100000
0.000342826
Dicing
Tag Test
0.000650941
Assembly Costs
0.00010000
1.00000000
0.00005651
0.00001000
Thinning
0.00002772
Dicing
0.10000000
Unit Machine
Line Max imized
$ Cost Pe r Die
$ Cos t Pe r Die
Thinning
0.01000000
$0.00142614
0.00100000
Flip Chip
Assembly
Tag Test
0.024655643
0.026883013
0.01857550
0.01959650
0.000342826
0.00083171
0.00010000
0.00001000
0.00002772
Line Max imized
Unit Machine
Worldwide Fab Capacity
z
z
z
z
Assume 1 billion
tags a day
100,000 dies/wafer
20% of world
silicon capacity
Fabs 15% idle
today
Outline, Part II
z
z
RFID and the Auto-ID Center
An in-depth look at some issues
z
z
z
z
z
A peek at the protocol
Security and Privacy Issues
Software
Vibration analysis
Silicon manufacturing
Conclusions
Physics and God
The Limits
Parameter #2
God
The limits of physics
vendors
Parameter #1