Applications of ICT to wildlife preservation efforts in India

Prabhat Ranjan

(prabhat_ranjan@daiict.ac.in)

Dhirubhai Ambani Institute of Information and Communication Technology,

Gandhinagar – 382007 (India)

Abstract

With the development of society, man and nature conflict is increasing day by day.

Areas traditionally reserved for wildlife are being encroached upon by human being.

This results in competition for same resource by human settlers and wildlife.

Developments in Information and Communication Technology (ICT) can allow one to

monitor the behaviour of wildlife in a way not possible earlier through traditional means

and help reducing chances of conflict. In this paper we discuss our work on using

sensor network technology to monitor wildlife in ways, not possible earlier. This work is

being done as a collborative research with Wildlife Institute of India, Deharadun(WII).

1. Introduction

With the development of ICT technology an area of research called “Sensor Networks”

has come up. This has been possible due to miniaturization of components,

development of low cost and low power integrated circuits, MEMS based sensor and

efficient wireless communication. Most of these developments have taken place due to

commercial demands but have led to sensor networks being a viable solution in many
application areas.

Now it is possible to integrate intelligent power saving devices on the body of the animal

which can monitor not only the migration pattern and activity but also the microclimate

through which animal is moving. All this data becomes available to the researchers at

their workplace through the wireless communication link. Similarly many more

possibilities have opened up making it easier for wildlife researcher to collect animal

behaviour in an efficient manner and improve wildlife conservation effort along with

developmental activities.

We describe here some of the project we are doing to utilize ICT in wildlife research by

the use of Sensor Network technology. We are working on three projects : (1) GPS

based system for tracking group of medium to large size animal behaviour, (2) GPS-

less small animal tracking system in Wildlife Institute of India, (WII), Deharadun and (3)

An image sensor network to monitor animal moving through a trail (such as Lion and

Tiger).

2. GPS based tracking system

Migratory pattern of wild animals is of paramount importance to wildlife researchers.

Traditionally human observers have played a important role in this. With the

development of ICT, GPS and satellite based radio collars have come into exixtence.

However, these are very expensive and due to this number of animals that can be

monitored using this technique is very small.

We are developing a sensor network based system using only terrestial peer-to-peer

communication for the purpose of monitoring movement of animal. We have tested a

proof-of-concept device [ 1 ], which can be put on the neck in the form of collar. This

device has a small embedded computer along with GPS and wireless communication

device. In addition device can be equipped with many sensors, depending on the

parameters we wish to monitor. For example, we have been working with monitoring

microclimate (Ambient Temperature, Humidity and Light) and animal head movements.

Information recorded using these sensors can be used to understand the migratory

behaviour much better. This information is not available in a similar work called

“Zebranet” [ 2 ]. The movement of the animal (using GPS) and other sensor data is

periodically recorded and stored on the device in a flash memory.

The stored data is spread to other animals carrying compatible nodes through peer-to-

peer commnunication. If any of the animals come in communication range with a fixed

or mobile base station, it transfers all the data it has accumulated through its own

measurements, or any other data through indirect/direct exchange with other animals.

In effect this means that wildlife researchers keep getting information about the

movement and other data of the group of animal at their convenience.

Hardware Details: The device on the body of animal would consist of an ATMEL

ATMega128L microcontroller, which is connected

to Lassen IQ GPS Receiver with embedded antenna for

collecting position information periodically. It is

connected to a 2.4 GHz radio transceiver which

allows wireless communication with other devices

or base station. A flash memory (AT45DB321B) of Figure 1 : GPS based tracking  

32 Mbits is used to store data being collected. device
Microclimate information would be collected through Temperature, Humidity (Sensirion

SHT11) and Light sensor (TAOS TSL2561). We are also working on adding an

accelrometer to monitor the head movement to find the activity animal is engaged in.

The device is powered through a battery with solar photovoltaic films to recharge them if

sunlight is available.

Device would be running an operating system to schedule its various tasks. Most of the

time device would be sleeping. It would have a RTC to wake it up periodically. Device

can be programmed to collect GPS and other sensor data at periodic interval. It would

also periodically check for the presence of other animals in the communication range

and perform exchnage of data. We have also provided for through the air configuration

of device for changing various parameters. Timing information obtained through GPS

would be used to synchronize the RTC and the various devices to each other.

3. GPS-less tracking of small animals

The method being developed above based on GPS measurement of position is not

suitable for small animals. Locating position using GPS consumes too much power,

resulting in battery weight to be much larger than desirable for a small size animal,

where one may have to limit device size on the body of the animal to be in the range of

25-30 gms. Thus one has to devise a new way of monitoring small animals over a

limited area like WII campus.

One way of doing this is to put a RFID tag on the animals and identify them when they

move in the range of a RFID tag reader. However the range of RFID reader is very

limited and thus requires very dense deployment of readers. This method is used for

animal, which moves in a well defined trail but is not suitable for small animals moving
in a wide area. Also it suffers from the fact that no microclimate (or other parameters

obtainable) can be recorded and studied. Some of the long range RFID systems have

been designed for identifying vehicles etc. with much larger weight than permissible for

monitoring small animals.

What we propose to try is to set up a

grid of transmitting/receiving stations

and put an active node on the body of

the animal. This can be seen in figure 2.

We have shown a representative set of

receiving stations (P1 to P9). This forms

4 different cells. Depending on the

location of the animal the four of the

stations would be communicating with Figure 2 : GPS­less tracking system for small  

animals. 'A' is the animal and P1..P9 is  
the animal (shown as 'A' in figure). The receiving stations.

device on the body of the animal would

be able to sense microclimatic parameters as well as transmit signals to communicate

with the receiving stations. We are assuming here that the receiving stations would

have power supply so we are not worrying about the energy consumption issues on the

receiving stations. However the device on the body of the animal would have severe

energy constraints and challenge of this work would be to maximize the life time of the

device without having to change the battery.

Now we have possibilities of two methods of localizing the animal. One is based on the

distance estimates obtained through signal strength received by the the neighbouring

receiving stations. We would carry out a mapping of signal strength at various locations
for a given transmitter. This mapping would be used to invert the signal strength

measurement to a location with respect to the grid points. [ 3 ] The second method

would depend on measuring of the angle the animal makes with respect to the various

receiving poles. This angle would be used to localize the animal. To find the angle we

would need to set up a scanning directional antenna on each of the poles. We are yet to

decide whether the scanning antenna would be motorized or based on electronic

switching of a multi-antenna system.

In addition, we would be recording microclimate (temperature, humidity and light)

through which the animal moves and would transmit this information to the receiving

station, once the communication is established. A critical part of the information would

be to determine when the animal is hybernating and for what period of time it remains in

that condition.

Hardware Details : The device on the body of the animal would consist of a Atmel AVR

microcontroller, a 2.4 GHz Zigbee transceiver, Flash memory, microclimate sensor and

battery. This would be supported by a RTC (Real Time Clock with battery backup). We

may also put a solar film for recharging the device and keep the battery weight low but

this will depend on whether animal comes out in the sun or not. Periodically, device

would try to communicate with the receiving stations. Period would be decided by the

average speed of the animal and residual energy in the battery.

The receiving stations will form a grid (not necessarily uniform as the terrain is not

uniform). Each station would have a Atmel AVR microcontroller, a flash storage unit and

a 2.4 GHz RF transceiver with appropriate set of antenna. In addition some of these

would also monitor local microclimate data. These nodes would be powered by campus

street light power supply and would not have a energy constraint like that of the device
on animal. The transceiver would always be in a listening mode. These devices would

form a multihop network and would eventually link to the Wi-Fi network of the WII

campus and would be able to transmit data collected to a server.

4. Image sensor network

Traditionally movement of animals such as tiger/lion which move through well defined

trail is done using camera placed at strategic locations. Currently WII uses film based

camera with PIR detector, which is placed across the trail. As the animal passses by the

camera is triggered and takes a picture. The cameras are placed in the evening and by

morning they are collected. A large number of cameras are used for this purpose and in

addition to it being cumbersome to deploy and collect it back, they are also prone to

theft.

We are developing a set of cameras which would get triggered by animals passing by

and stored picture would be transmited through wireless link. To start with we will be

using a GPRS modem to transmit the picture

in the WII campus but later on we would

connect a Zigbee transceiver and use a

Zigbee/GPRS gateway to transmit the picture

as mobile signals would not be available in

the forest area. The cameras would form

networks to make optimal utilization of the

network at the same time they would keep

track of the animals movement to prepare Figure 3 : Camera device for tracking  

animals on trail
themselves better.
Hardware Details: We will be using an image sensor module with digital output and

compression capability based on OmniVision OV640/8 VGA Color Digital image sensor

with OV528 JPEG compression chip. This module can take images both video and still

upto 640x480 resolution. This will be combined with KC7783 PIR Detector Module to

detect the motion of the animal upto 10 feet away to trigger the camera. The device

would be integrated with an ATMega128 microcontroller and a GPRS modem to

transmit pictures through mobile network to base computer. Figure 3 shows the various

components for this device.

We will also consider the possibility of splitting the system in two parts for reducing the

size of camera device, which can be placed near to the trail and the other part for long

range communication little farther off.

5. Summary

We have discussed ICT projects that we are working on right now in collaboration with

Wildlife Institute of India, Deharadun. We feel that success of these projects would open

up a new chapter of cooperation between ICT professionals and wildlife researchers all

over country. This would pave the way for many more projects, where ICT can be used

to improve the socio-economic conditions.

6. Acknowledgements

We would like to acknowledge Wildlife Institute of India, Deharadun which is funding

these projects as a collaborative research. We would like to thank Mr P R Sinha

(Director, WII) and Dr. V B Mathur (Dean, WII) for taking great interest in these projects
and making sure that necessary apporvals are done quickly. We have had number of

dicussions with scientists in WII to finalize the goals of these projects and we would

specially like to thank Dr. S P Goyal, Dr. Bivash Pandav and Mr Qamar Qureshi for

extensive discussion during my visits to WII and through other channels of

communication.

Device development for these projects is being done with Prof. Rahul Dubey. Work on

proof-of-concept device was done by Prabhat Saraswat, Ashish Kumar, Swetha Polana

and Amit Singh as part of their BTech final semester project and Udayan Kumar

(Research Engineer).

7. References

1. P. Ranjan, P. Saraswat, A. Kumar, S. Polana and A. Singh, “wildCENSE :

Sensor Network for wildlife monitoring”, May 2006.

( http://intranet.daiict.ac.in/~ranjan/research/papers/wildCENSE_REPORT.pdf)

2. P. Zhang, C. Sadler, S. Lyon, and M. Martonosi, “Hardware Design Experiences

in ZebraNet,” Proceedings of the ACM Conference on Embedded Networked

Sensor Systems (SenSys) 2004, November 2004.

3. Nirupama Bulusu, John Heidemann and Deborah Estrin, "GPS-less Low Cost

Outdoor Localization For Very Small Devices", IEEE Wireless Communications,

Vol 7. No.5, pp. 27-34, Oct 2000.