In-Pavement Wireless Sensor Network for Vehicle Classification Ravneet Bajwa, Ram Rajagopal, Pravin Varaiya and Robert Kavaler IPSN’11

In-Pavement Wireless Sensor Network for Vehicle

ClassificationRavneet Bajwa, Ram Rajagopal, Pravin Varaiya and

Robert Kavaler

IPSN’11

Outline

• Motivation• Introduction• Description• Communication Protocol Design• Experiment Setup• Performance• Conclusion & Future Work

Outline


Motivation

• Intrusive technologies– Piezoelectric sensors, inductive loops– High installation and maintenance costs

• Non-intrusive technologies– Infrared, video imaging– Sensitive to traffic and weather condition

• Propose an alternative system base on a WSN that is both cost effective and insensitive to environmental conditions


Problem Statement

• Cars, buses, three-axle single unit trucks, and five-axle single trailer trucks

• A vehicle travels in a traffic lane at some varying speed and we wish to count the number of axles and the spacing between each axle in an accurate manner

Proposed WSN System

• Vibration sensor (accelerometer) embedded in the road– Calculate the axle spacings

• Vehicle detection sensor (magnetometers)– Report the arrival and departure times of a vehicle

• Access point (AP)– Send commands to sensors– Log the incoming data

• First in-pavement, easyto deploy, WSN basedsystem for counting axlesand axle spacing

Outline


Wireless Vehicle Detection Sensor

• Measures the changes in magnetic field to infer the local presence of a vehicle

• Synchronous Nanopower Protocol(SNP), aTDMA based protocol– Last 10 years with a single 7200 mAhr battery

• Given the arrival times tai and taj at the twosensors i and j, the speed v will bev = dij / |taj – tai|

• Estimate the length(L) of the vehicleL = v(tdj - taj)

Wireless Vibration Sensor

• Sample the analog output of an accelerometer and transmit the data via a radio

• Sample fast enough to capture the transient vibrations• Sensor needs to be insensitive to the vehicles traveling in the

neighboring lanes• Insensitive to the truck engine and environmental noise• Sensor resolution target is 500 ug• Bandwidth 50Hz• Sampling frequency 512 Hz( > 5 times Nyquist Frequency)

– Power consumption increases for higher sampling rates

Selecting an accelerometer

• SD1221-005 has higher sensitivity and lower noise density

• However, it consumes more than 20 times the current than MS9002.D and has to be operated at higher voltage

• Both devices achieved the aimed minimum resolution of 500 ug– Select MS9002.D due to its low operating voltage and low

current consumption

Filters for mitigating sound noise

• Accelerometer is sensitive to sound• MS9002.D behaves like a microphone under the

device’s bandwidth• 3rd order low-pass filter with cutoff frequency of 50

Hz is sufficiently aggressive to filter out most of the sound in the audible spectrum

Casing

• Sound isolation• Protect the electronics from

rain water and oil spill on theroad

Circuit Description

• 2.5 V supply voltage• Amplifier with gain 10• The gain of 10 reduces the range of the

accelerometer to ≈±225mg• This is necessary in order to ensure

that the quantization noise from the ADC is less than the noise from the accelerometer– Otherwise, the resolution of the system will be limited by ADC noise

• The reduced range is still sufficient– For heavy trucks ± 200 mg

Outlin


Communication Protocol Design

• MAC Layer– TDMA based– Time is divided into multiple frames with each frames

about 125 ms long– Each frame is further divided into 64 time slots– Slot 0 is used by AP to send clock synchronization

information and other commands to the sensors– AP assigns every node unique time slots and a node ID to

communicate with it.

Application Layer

• Sync Application– AP sends sync packets on a periodic basis– Sensor node listens to sync packets every 125 ms– When the clock converges to steady state, then is listens

for a sync packet only once in 30 s– Sync application is also used to send commands– Set Mode, Reset, Set Timeslot, Set RF, Download Firmware,

Set ID

Application Layer

• Accelerometer Application– Idle Mode: accelerometer and related circuitry are

turned off by disabling the voltage regulator• Once every 30 s, the microcontroller and the

transceiver wake up and acquire the sync packet

Application Layer

– Raw Data Mode: microcontroller wake up every 1/512 s, and samples the analog output from accelerometer• 32 samples at a sampling freq. 512Hz, and each sample

containing 12 bits of information• In every frame(125ms) we accumulate 96 bytes of

information to transmit• To have a reasonable packet size, we fragment the data

in two parts, 48 bytes each, and transmit it using two different time slots 62.5ms apart

Application Layer

• Download Firmware Application– Reprogram the entire flash memory of a sensor node over

the air– AP transmits new code repeatedly and the node updating

its code in small pieces– Only the data that do not overwrite the current running

program are updated by the node

Axle Detection(ADET) Algorithm

Axle Detection(ADET) Algorithm

• Using data from 4 trucks at different speeds, we observed the bandwidth of the energy signal and empirically defined by M(v) = 900/v

• Low-pass filter is optional• Minimum time separation ζ(v) was chosen by

assuming that the axles are at least 6ft apart

Wide Lane ADET Algorithm

• Wander movement in a lane• Combining vibration readings from multiple sensors• Delay Di = di / v

Outline


Experiment Setup

• 4 vibration sensors and 4 vehicle detection sensor were installed on California Highway I-680

• Vehicles come from Sunol Weigh Station

• Slow down at weigh station– Easy to collect ground truth

• Data from 53 different trucks, rangingfrom pickup trucks to 5-axle commercial trucks

Installation

• Boring a 4-inch diameter hole approximately 2.25 inches deep• Installed on a road in less than 20 minutes• Installation of a small sensor is much cheaper and convenient

than installing special material pavements required for piezoelectric sensors

Deployment Challenges

• Packet Drops– Drop rate was low(1%) retransmit packets with a delay of

1 packet drop rate is almost 0• Packet 1, 2, 1, 2

• Vehicle Wander– use Wide Lane ADET algorithm

• Sensor failure– Sensor k did not work– Vibration data was available from 3 sensors

Outline


Vibration Sensor Performance

• Noise with no vehicle in vicinity– 414 ug RMS

• Truck was parked on top of the sensor with engine were on vs. truck blew its horn– 7% vs. 4%

• With a heavy truck traveled in the closed lane– Sensor did not register any noticeable peaks

Axle Count

• Error difference between the ground truth axle count and the estimated axle count

• By combining the measurements from all sensors, the algorithm always gives the correct axle count

• Error results form the wander movement

Axle Spacing

• Left: for tandem axle• Middle: pick up trucks, small two axle

commercial trucks• Right: axles of trailers

Outline


Conclusion

• A novel algorithm that estimates the axle count and spacing from pavement acceleration was designed and tested on the collected data

• ADET is simple enough to implement a sensor node with limited processing power

• Majorities of the existing technologies are wired solutions • Both the sensors and the AP are powered by batteries and

consume much less power than other technologies• The installation procedure and sensors themselves are much

cheaper• There is minimal maintenance compared to other technologies

Future Work

• Find an optimal arrangement of sensors in order to minimize the number of sensors deployed

• Reduce the amount of data transmitted• Reduce the sensor power consumption

Documents

In-Pavement Wireless Sensor Network for Vehicle Classification Ravneet Bajwa, Ram Rajagopal, Pravin Varaiya and Robert Kavaler IPSN’11