GDI 2003: status report

1. GDI 2003: status report Robert Szewczyk Joe Polastre Alan Mainwaring David Culler NEST Retreat, Jan 15, 2004

2. Outline • GDI • Mote design • Networking improvements • Infrastructure redesign • Conclusions Analysis Design Deployment

3. Scientific motivation: Leach’s Storm Petrel • Questions • What environmental factors make for a good nest? How much can they vary? • What are the occupancy patterns during incubation? • What environmental changes occurs in the burrows and their vicinity during the breeding season? • Methodology • Characterize the climate inside and outsize the burrow • Collect detailed occupancy data from a number of occupied and empty nest • Spatial sampling of habitat – sampling rate driven by biologically interesting phenomena, non-uniform patches • Validate a sample of sensor data with a different sensing modality • Augmented the sensor data with deployment notes (e.g. burrow depth, soil consistency, vegetation data) • Try to answer the questions based on analysis of the entire data set

4. Patch Network Sensor Node Sensor Patch Gateway Transit Network Client Data Browsing and Processing Basestation Base-Remote Link Internet Data Service Application architecture

5. Sensor node evolution

6. Sensor node GDI ’02 • Mica platform • Atmel AVR w/ 512kB Flash • 916MHz 40kbps RFM Radio • Range: max 100 ft • Affected by obstacles, RF propogation • 2 AA Batteries, boost converter • Mica weather board – “one size fits all” • Digital Sensor Interface to Mica • Onboard ADC sampling analog photo, humidity and passive IR sensors • Digital temperature and pressure sensors • Designed for Low Power Operation • Individual digital switch for each sensor • Designed to Coexist with Other Sensor Boards • Hardware “enable” protocol to obtain exclusive access to connector resources • Packaging • Conformal sealant + acrylic tube

7. GDI ’02 population • 43 distinct nodes reporting data between July 13 and November 18 • Heavy daily losses • Between 3 and 5% daily

8. Redesign directions • Node-level issues that need resolving • Size – motes were too large to fit in many burrows • Application specific packaging; minimize size of burrow package, base the system around mica2dot • Packaging – did not provide adequate protection for electronics or proper conditions for sensors • Waterproof plastic packaging to protect electronics • Design to provide both shielding and exposure to sensors • Node reliability • Power consumption – boost converter not particularly useful • Eliminate boost whenever possible, use stable voltage lithium cells • Data interpretation challenges • Sensor calibration • Occupancy data interpretation – need more sophisticated processing of sensor data and/or ground truth data • Better metadata – sensor location & conditions

9. Miniature weather station • Sensor suite • Sensirion humidity + temperature sensor • Intersema pressure + temperature sensor • TAOS total solar radiation sensor • Hamamatsu PAR sensor • Radiation sensors measure both direct and diffuse radiation • Power supply • SAFT LiS02 battery, ~1 Ah @ 2.8V • Packaging • HDPE tube with coated sensor boards on both ends of the tube • Additional PVC skirt to provide extra shade and protection against the rain

10. Burrow occupancy detector • Sensor suite • Sensirion humidity + temperature sensor • Melexis passive IR sensor + conditioning circuitry • Power supply • GreatBatch lithium thionyl chloride 1 Ah battery • Maxim 5V boost converter for Melexis circuitry • Packaging • Sealed HDPE tube, emphasis on small size

11. GDI ‘03 weather mote population

12. GDI ‘03 burrow mote population

13. Patch network GDI ‘02 • Single hop transmit-only network • 43 nodes, about 25 above ground, the rest in burrows • Repeater network – add an extra hop to improve connectivity into burrows • Ran out of energy before it made any difference • Sampling rates: 1 set of samples from every node every 70 seconds • A compromise between response time (and ease of deployment) and expected power management behavior • Application logic: sense (fix time), send, sleeep(fix time) • Expected that CSMA MAC backoff will effectively desynchronize all nodes

14. GDI ’02 deployment

15. Software architecture advances • Bi-directional communication with low-power listenting • 1% duty cycle • Parameter adjustment and query • Sample rate changes, sensor status queries • Improved power management scheme • Fine granularity through StdControl interface • 10 uA sleep mode, 30 uA with running Timer.

16. GDI ’03 patch network • Single hop network deployed mid-June • Rationale: Build a simple, reliable network that allows • HW platform evaluation • Low power system evaluation • Comparisons with the GDI ’02 deployment • A set of readings from every mote every 5 minutes • 23 weather station motes, 26 burrow motes • Placement for connectivity • Network diameter 70 meters • Asymmetric, bi-directional communication with low power listening – send data packets with short preambles, receive packets with long preambles • Expected life time – 4+ months • Weather stations perform considerably better than burrow motes – their battery rated for a higher discharge current

17. Packet yields

18. GDI ’03 deployment

19. GDI ’03 Multihop network • Motivation • Greater spatial reach • Better connectivity into burrows • Implementation • Alec Woo’s generic multihop subsystem • Low power listening: tradeoff channel capacity for average power consumption • Contrast with TASK approach: Alec’s multihop component but with duty cycling on a loosely synchronized network • The network nodes • 44 weather motes deployed July 17 • 48 burrow motes deployed August 6 • Network diameter – 1/5 mile • Duty cycle – 2% to minimize the active time (compromise between receive time and send time) • Reading sent to base station every 20 minutes, route updates every 20 minutes. Expected lifetime: 2.5 months • 2/3 of nodes join within 10 minutes of deployment, remainder within 6 hours. Paths stabilize within 24 hours

20. Packet Yield

21. Topology stability

22. Parent-child link distribution

23. Parent-child link longevity distribution

24. GDI ’02 base station: • Requirements • Disconnected operation • Remote management • Automatic restart • Redundancy • Implementation • 2 laptops, each with a direct serial connection to a transit network (via GenericBase) • Asymmetry: one of the laptops acting as a gateway/firewall • Limited inside network • Replicated but independent PostgreSQL servers provide resiliency against laptop crashes • Limited remote admin capability – remote desktop, ssh • How do you reboot a system 3000 miles away • Satellite connection • DirecWay WAN • Uptime: 47%

25. Web power strip Satellite router WWW power strip DB 4-port VPN router and 16-port Ethernet switch IBM laptop #1 DB Power over LAN midspan IBM laptop #2 Wireless bridge Axis 2130 PTZ South Southern WAP Mica2-EPRB#2 Mica2-EPRB#2 GDI ’03 Base Station • More sophisticated networking structure • Dual laptops with PostgreSQL • Dual base stations (Mica2 + EPRB) • But one logs single hop the other logs multihop • Cross logging of the data • Vastly improved remote access • Remote wakeonlan • Web enabled power strip • Ubiquitous POE • VPN for direct access from authorized networks • Extensible schema to accommodate new sensor modalities and query types, compatibility with TASK • Main stumbling block • Power, power, power • Lack of redundancy on the transit net • Minor HW issues – outdoors is harsh

26. Occupancy measurements GDI ‘02

27. Wireless bridge Burrow Camera Configuration 12V PoL Active Splitter 12VDC, 0.9A network Axis 2401 Video Server Sensor Patch IR Burrow Camera #5 IR Burrow Camera #1 IR Burrow Camera #6 IR Burrow Camera #2 IR Burrow Camera #7 IR Burrow Camera #3 ) Power over LAN Midspan Ethernet switch IR Burrow Camera #4 IR Burrow Camera #8 110VAC service Northern WAP Occupancy measurements GDI ‘03 • Calibrated ASIC for conditioning and processing the passive IR signal • 0 to 40 deg C range • Corroboration of data • Multiple sensor nodes in occupied burrows • Verification of data • Co-locate a completely different sensing network with motes • IR-illuminated cameras • Ethernet video servers • Wireless connection to the base station • Verification network mimicsthe architecture of the sensornet • Sample a 15 sec video/audio clipevery 5 minutes • ~6 GB worth of data so far…

28. Occupancy data evaluation status • PIR data from website used for finding occupied burrows • Saturated sensor outputs • Video data analysis underway • Entry/exit events • Automatic video analysis

29. Analysis from biology side Temperature and humidity distributions

30. Temperature and humidity in different habitats Weather stations Burrow motes

31. Climate data Weather stations Burrow motes

32. Climate data: day-to-day variations,meadow Weather stations Burrow motes

33. Conclusions • Another iteration on the design, deploy, analyze cycle • 50+ node single hop network, 100+ node multi hop network • 4.5 months of operation – June 8 – October 20 • 436 thousands weather station observations • 234 thousands burrow mote observations • Improvements in the network quality • Longevity, reliability, features, power management, data quality • Room for much more improvement – ease of packaging, robustness • Data analysis • Biologists are engaged • Video analysis for occupancy corroboration under way

34. Q&A Thank You! http://www.greatduckisland.net

35. Gateway node design • GDI ’02 implementations • Linux + CerfCube + 802.11b • Generic Base x 2 + omni-directional antenna to receive from patch + directional antenna to xmit to base station • Power requirements: • CerfCube = 30”x30” solar • Mote-based = 6”x6” solar • Reliability • Side effects: transit network transmissions will affect the transmissions in the patch

36. Gateway node design • GDI ’03 design – keep elements that worked • Keep a mote-based system • Use different frequencies on patch and transit networks to eliminate interference; different frequencies for single and multihop networks • Asymmetrical bi-directional communication on single hop network – exploit low power listening & always-on gateways • Symmetrical bi-directional communication in the multihop network • Storage and processing • Keep it simple • No storage, processing, aggregation, etc. • Big disadvantages – removes a layer that could buffer packets in case of transient failure elsewhere in the system (e.g. base station down)