GDI 2003: status report

Robert SzewczykJoe Polastre

Alan Mainwaring David Culler

NEST Retreat, Jan 15, 2004

Outline

• GDI

• Mote design

• Networking improvements

• Infrastructure redesign

• Conclusions

Design

Deployment

Analysis

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

Application architecture

Base-Remote Link

Data Service

Internet

Client Data Browsingand Processing

Transit Network

Basestation

Gateway

Sensor Patch

Patch Network

Sensor Node

Sensor node evolution

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

GDI ’02 population

• 43 distinct nodes reporting data between July 13 and November 18

• Heavy daily losses– Between 3 and 5% daily

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

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

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

GDI ‘03 weather mote population

Jun Jul Aug Sep Oct Nov0

5

10

15

20

25

30

35

40

45

No

de

cou

nt

Multihop weather mote population

Jun Jul Aug Sep Oct Nov0

5

10

15

20

25

30

No

de

cou

nt

Singlehop weather mote population

GDI ‘03 burrow mote population

Jun Jul Aug Sep Oct Nov0

5

10

15

20

25

30

No

de

cou

nt

Single hop burrow mote population

Jun Jul Aug Sep Oct Nov0

10

20

30

40

50

60

70Multihop burrow mote population

No

de

cou

nt

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

GDI ’02 deployment

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.

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

Packet yields

GDI ’03 deployment

10m

Single hop weatherSingle hop burrowMulti hop weatherMulti hop burrow

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

Packet Yield

Topology stability

0 20 40 60 80 100 1200

500

1000

1500

2000

2500

3000

3500

Mote ID, ordered by number of received packets

Pac

ket

s re

ceiv

ed f

rom

a m

ote

Distribution of received packets. Color patches correspond to a particular parent.

Average messages per mote: 1055Average number of parents: 6.7

Parent-child link distribution

-0.2 0 0.2 0.4 0.6 0.8 1 1.20

0.1

0.2

0.3

0.4

0.5

0.6

0.7Distribution of parent-child relationship

Relationship fraction

Example of relationship fraction:

10% of packets from mote X have been

delivered through mote Y

Fra

ctio

n o

f lin

ks

Parent-child link longevity distribution

0 50 100 150 200 250 300 350 400 450 5000

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

Fra

ctio

n o

f lin

ks

Messages delivered over the link

Average messages delivered over a link: 158

Link longetivity distribution

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%

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 harshMica2-EPRB#2

IBM laptop #1DB

Web power strip

Axis 2130 PTZ South

Wireless bridge

4-port VPN router and

16-port Ethernet switch

Power over LAN midspan DB

IBM laptop #2

Mica2-EPRB#2

WWW power strip

Southern WAP

Satellite router

Occupancy measurements GDI ‘02

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 mimics

the architecture of the sensornet

– Sample a 15 sec video/audio clipevery 5 minutes

– ~6 GB worth of data so far…

Sensor Patch

Power over LAN Midspan

IR Burrow Camera #1

IR Burrow Camera #2

IR Burrow Camera #3

)IR Burrow Camera #4

IR Burrow Camera #5

IR Burrow Camera #6

IR Burrow Camera #7

IR Burrow Camera #8

Axis 2401 Video Server

12VDC, 0.9Anetwork

Burrow Camera Configuration

Northern WAPEthernet switchWireless bridge

12V PoLActive Splitter

110VAC service

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

12:00 13:00 14:00 15:00 16:00 17:0039

39.5

40

40.5

41

Time

PIR

Tem

per

atu

re (

C)

PIR and ambient temperature, Burrow 3, Aug. 4

12

13

14

15

16

Am

bie

nt

Tem

per

atu

re (

C)

00:00 01:00 02:00 03:00 04:00 05:00 06:0010

15

20

25PIR and ambient temperature, Burrow 3, Jul 13

PIR

Tem

per

atu

re (

C)

10

15

20

25

Am

bie

nt

Tem

per

atu

re (

C)

Analysis from biology side

Temperature and humidity distributions

Temperature and humidity in different habitats

M E WHabitat

0

7

14

21

28

35

Mea

n Te

mpe

ratu

re (

d egr

ees

C)

M E WHabitat

0

7

14

21

28

35

Mea

n Te

mpe

rat u

re (d

egre

es C

)

Weather stations Burrow motes

165 184 203 222 241 260Day of Year

0

10

20

30

Mea

n Te

mpe

rat u

re (

d egr

ees

C)

Climate data

165 184 203 222 241 260Day of Year

0

10

20

30

Mea

n Te

mpe

rat u

re (

d egr

ees

C)

Weather stations Burrow motes

Climate data: day-to-day variations,meadow

160170180190200210220230240250260Day of Year

0

10

20

30

Mea

n Te

mpe

ratu

re (

degr

ees

C)

140 160 180 200 220 240 260Day of Year

0

10

20

30

Mea

n Te

mpe

rat u

re (

d egr

ees

C)

Weather stations Burrow motes

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

Q&A

Thank You!

http://www.greatduckisland.net

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

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)