Sphere of Influence Project owner: Mark Yarvis Intel
Download
Report
Transcript Sphere of Influence Project owner: Mark Yarvis Intel
Copyright© Intel Corporation 2000-2004
Industrial Applications
for Sensor Networks
Condition based monitoring pilot project
Lama Nachman
Lakshman Krishnamurthy
Researcher
Hans Mulder
Intel Research and System
Technology Lab
Ralph Kling
Mark Yarvis
Jasmeet Chhabra
Carl Dellar
®
Copyright© Intel Corporation 2000-2004
Agenda
Introduction & Problem Statement
Equipment Health monitoring Pilot in Intel FAB
Application requirements
Current Implementation (Phase 2)
MICA & iMote clusters
Reliability protocol
Network Configuration
Power Saving Protocol
Status & Next steps
Key learnings
Intel Research
•2•
Copyright© Intel Corporation 2000-2004
Fab Pre-emptive Maintenance
Application at Intel
Use vibration signatures to
identify problems with equipment
Avoid failure
~5000 Sensor points in each fab
4 years of archived data
Done by sneaker net today
Move to wireless sensor network
Demonstrate a commercially feasible ROI for sensor
and mesh network deployments
Intel Research
•3•
Copyright© Intel Corporation 2000-2004
Problem Statement
Equipment failures in live production fabs is
extremely costly ($Millions)
Shutdown results in opportunity loss
Cost of evacuation and requalifying all the tools
Possible loss of wafer lots in the pipeline
Need to predict equipment failures early
enough and perform preemptive maintenance
during pre-scheduled down-time
Monitor equipment health using vibration
signatures
Intel Research
•4•
Copyright© Intel Corporation 2000-2004
Case study (RA FAB)
~5000 sensing points already instrumented
40% permanent sensors, 60% portable sensors
Vibration and RPM sensors (Wilcoxon & Honeywell)
Manual data collection using handheld devices
Time domain data is collected, spectrum and magnitude
plots are generated
Data is downloaded to Rockwell Enshare software
Sensors are manually configured in DB
Type, location, direction, collection frequency, etc
Alarms are generated, further manual collection is
performed on specific sensors
Intel Research
•5•
Copyright© Intel Corporation 2000-2004
Case Study (RA FAB)
Prevention estimates
Once per month -> catch (80-85)%
Once per week, and selective daily collections -> catch
~97%
Manual collection method is currently used
Target is once per month
Headcount cost : ~$500,000 in one FAB
Rockwell based solution (EnWatch)
Ethernet based on-line system (~$5000)
16 channels, data collection and analysis
Controlled by EnShare backend
Intel Research
•6•
Copyright© Intel Corporation 2000-2004
Application Requirements
Interface to Wilcoxon vibration sensors and Honeywell RPM
sensors
0.5 Hz – 5KHz range
3000 Samples, 16 bits each
Collect once per week (optional selective collection)
Battery life
No access to power or Ethernet at sensing locations
6 months @ 1 collection per month
4 months @ 1 collection per week
Reliability
MTBF : 6 months
Identify bad data (especially false good data)
Interface to Rockwell EnShare backend
Automatic network configuration and maintenance
Intel Research
•7•
Copyright© Intel Corporation 2000-2004
Pilot Network Architecture
Fab Equipment
Intranet
Intranet isolation
Root Node
Ad Hoc Mote
Network
Cluster Heads
Mote + Vibration Sensors
Intel Research
•8•
Copyright© Intel Corporation 2000-2004
Solution components
Ad hoc Mote network
MICA based clusters
Imote based clusters
End to End Reliable datagram transport protocol
(sensor node -> Root Node)
802.11 overlay mesh network using stargates
Cluster head manages data collection and
sleep/wake schedule
Root Node collects the raw data, stores in EnShare
format and sends it to server
EnShare data base imports the raw data
Intel Research
•9•
Copyright© Intel Corporation 2000-2004
Reliability Protocol
Runs on Mica motes, iMotes and Stargates
TinyOS implementation
Provides VarSend, VarRecv interfaces to app layer
Uses Generic Packet interface to abstract network layer
Sliding window protocol
Connection parameter negotiation (fragment size,
window size, timeout info)
Receiver sends an ACK bitmap within window
Sender retransmits NACK’d fragments
3 phases
Connection setup (light weight, 2 packets)
Data exchange (data and NACK packets)
Final ACK (2 packets)
Intel Research
• 10 •
Copyright© Intel Corporation 2000-2004
Mote Cluster
Implementation
Intel Research
• 11 •
Copyright© Intel Corporation 2000-2004
Data Collection / Power Saving
Cluster head sends a command to each
sensor node to start data collection
Sensor node initiates reliable transport
protocol with RootNode for each connected
sensor
Sensor node informs cluster head when data
transfer is complete
Cluster repeats the process for each sensor
node
When all sensors have been collected, the
complete cluster is put to sleep until next
collection
• 12 •
Intel Research
Copyright© Intel Corporation 2000-2004
iMote Cluster Details
Intel Research
• 13 •
Copyright© Intel Corporation 2000-2004
Intel Mote: an enhanced wireless
network research platform
Intel Mote is a modular,
stackable design
Hardware features
High platform integration level (core,
radio, memory…)
Main board (ARM core, SRAM,
FLASH, BT radio)
Low power operation
Power supply board (battery,
AC, solar, …)
Small physical size
Modular HW/SW design
Sensor board(s)
Low cost and volume production
potential
Other boards (alternate radio,
debug, actuator, …)
TinyOS applications
TinyOS base components
Network layer (multihop)
Sensor board
Main board
Intel Mote layer
Power board
Firmware (BT-LLS)
Backbone interconnect
Hardware
Intel Research
• 14 •
Copyright© Intel Corporation 2000-2004
Network Configuration
Automatic scatternet formation algorithm
Forms a tree structure
Clusterhead is the root of the tree (Master Role)
Intermediate nodes have dual Master/Slave roles
Leaf nodes are slave only nodes
Free nodes alternate between BT Inquiry & scan
modes to discover other nodes
Free nodes can join at different levels in the tree,
depending on which node they connect to
Connected nodes only scan to eliminate the
possibility of creating loops
Simple routing algorithm
Intel Research
• 15 •
Copyright© Intel Corporation 2000-2004
Power Saving Protocol
Leverage low power modes in Bluetooth
Cluster head broadcasts a “network sleep” message
down the tree.
Once the message reaches a leaf node, a response
is sent up the tree
When a master hears a response from all its slaves,
it will put all the links on hold, and propagate the
response up the tree
Messages can still flow through the network in
between hold intervals (20 second response time
per level in the tree)
The cluster will broadcast a “network wake up”
message down the tree
Intel Research
• 16 •
Copyright© Intel Corporation 2000-2004
Network Observations
1 minute to form a cluster of 16 nodes
BT links are very stable once established
Network formation overhead is amortized
over long connection time
BT link layer reliability is very effective,
hence reducing the end to end NACKs
Need to optimize the scatternet formation
algorithm to select connections based on link
quality, and reducing hop count
Intel Research
• 17 •
Copyright© Intel Corporation 2000-2004
iMote Cluster integration
iMote cluster
Simple routing algorithm
iMote MHOP header (src, dest, channel)
TOS Message is not used
MICA cluster & RootNode
DSDV & flood protocols
iMote cluster head translates between domains
Route update messages from rootnode intercepted to get
the RootNode ID
Reliability protocol hdr/data is repackaged
Sensor -> RooNode (iMote packet -> DSDV packet)
RootNode -> Sensor (Flood -> iMote packet)
Intel Research
• 18 •
Copyright© Intel Corporation 2000-2004
Sensor Board
18V power supply
10kHz+ 24bit A/D
Programmable antialiasing filter
PLD bridges SPI to UART interface
Wilcoxon sensor
Voltage
output
Intel Research
Sensor
voltage
supply,
A/D, filter
• 19 •
Intel Mote
UART
900kb/s
SRAM 64kB
FLASH 512kB
Copyright© Intel Corporation 2000-2004
Time domain data
Intel Research
• 20 •
Copyright© Intel Corporation 2000-2004
Frequency domain data
80Hz
reference
signal
Intel Research
• 21 •
Copyright© Intel Corporation 2000-2004
Status
Phase 2 development complete
Testing will begin in the JF3 chiller room with
MICA & iMote clusters next week
Hardware is installed CUB3
CUB3 deployment is scheduled for mid June
Collecting performance and power data for
platform comparison by end of June
Intel Research
• 22 •
Copyright© Intel Corporation 2000-2004
JF3 Pilot Deployment
Facilities
Rooms
Intel Research
• 23 •
Copyright© Intel Corporation 2000-2004
Next Steps
Finalize phase 3 requirements
Choose one Mote platform based on the
phase 2 data
Move to TinyDB/TASK
Intel Research
• 24 •
Copyright© Intel Corporation 2000-2004
Key learnings (platform)
Size requirements
Not very sensitive
Current solutions are much larger
Mote size is negligible
Power Consumption
Sensor + A/D consume a lot (~60 mW)
Can use large batteries
Large RAM is very useful
Adding more capabilities to the mote simplifies the
sensor board design
Fast I/O on the mote is useful
Intel Research
• 25 •
Copyright© Intel Corporation 2000-2004
Key learnings (Network)
Automatic configuration of the network is required
Reducing hop count is key
Heterogeneous networks are very useful
Simple power saving protocols are sufficient
(Cluster based)
Matching radio bandwidth to application
requirements can save power
Fast network response time is needed, even if
collection frequency is infrequent
Polling specific sensors and adding streaming
modes will be very useful
Intel Research
• 26 •
Copyright© Intel Corporation 2000-2004
Key learnings (Network)
Debug modes would be very useful
Tracing network topology and data flows
Performance and power monitoring
Isolating bad data and recovering from
failures
Intel Research
• 27 •
Copyright© Intel Corporation 2000-2004
Key Learnings (backend)
Interfacing into existing tools is extremely
important
Want to use Rockwell EnShare for network
control/command
Getting info into/out of Rockwell was very painful
Easing the installation process is very
desirable
Automatically recognizing sensors and their
location is very useful (sensor -> equipment
mapping)
Intel Research
• 28 •