FOREST WATCH

1
FOREST WATCH nsitspirit
Team Members Manav Raman Manish Bhardwaj Pulkit
Midha Vamsi Narla Mentor Dr. M.P.S Bhatia
2
MOTIVATION
3

Motivation

• Annual losses from forest fires in India are
  estimated at Rs 440 crores (US 107 million).
  This estimate does not include the loss suffered
  in the form of loss of biodiversity, nutrient and
  soil moisture and other intangible benefits.
  (Ref Forest Survey of India)
• In United States, around 2.8 million ha of
  forests were destroyed by forest fires, compared
  to around 1.7 million ha in 2002.
• Indonesias forests loss was about 6.2 million
  acres a year.
• Globally, more than 350 million ha of forests
  were burned in 2000. This area equals the size of
  India.
• Ref (FAO , UN 2003)

4
Motivation.
Illegal Logging (In Indonesia) The forest
ministrys figure of production of timber
products 882 million cubic
feet Consumption of Timber related Products
2.6 billion cubic feet
-- -----------------------------------
------ Deficit -1.7
billion cubic feet This means that industries
obtains 73 of their logs from illegal or
unsustainable sources. Resulting in state
financial losses of approximately 3.37
billion. (Ref The Indonesia-United Kingdom
Tropical Forest Management Program survey, 2004
).
5
Impact of Forest Fires and Illegal logging

• Health problems leading to diseases.
• Loss of valuable timber resources and depletion
  of carbon sinks resulting in Global Warming.
• Change in micro climate of the area making it
  unhealthy living conditions.
• Soil erosion affecting productivity of soils and
  production
• Degradation of water catchments areas resulting
  in loss of water.
• Loss of biodiversity and extinction of plants and
  animals.
• Loss of natural regeneration and reduction in
  forest cover and production.
• Ozone layer depletion.

6
Performance Features

• Forest Watch is a Wireless Sensor Network based
  system that has the following performance
  features
• Real-time System
• Detects and Locates a Forest Fire
• Detects Illegal Logging
• Alarm-raising for above scenarios
• Validation of alarm by Tier-2 of the system
• Data Warehousing
• Fire Occurrence Prediction
• Detailed Monitoring and Mapping of Fire-Prone
  Zones
• Low Power Consumption
• Remotely Updatable

7
System Block
RF Link
Sensor Node
eBox
GSM Link
Central Workstation
8
System Implementation
9
Ebox polls each node one by one
Referenced node sends sensor data to Ebox
Ebox reads sensor data for each node, builds
database
10
Our Innovation

• A CMOS camera based validation system which
  verifies the occurrence of the fires as reported
• An Algorithm has been developed by us to
  determine the position of each node with respect
  to the eBox, eliminating the need for GPS.
• A Multi-Hop Routing Protocol has been developed
  by us for wireless sensor networks. The protocol
  is Fault Tolerant.
• It is specific to the topology of our project.
• Detection of illegal logging.

11
Comparison with Existing Systems

• The existing Satellite Imaging based fire alarm
  systems are not real-time and typically have a
  slow response time.
• Satellite systems are expensive as compared to
  Forest Watch.
• No existing system deals with the problem of
  illegal logging.
• It is capable of taking pre-emptive action on the
  basis of patterns of previously recorded
  temperature data, before and during forest fires
  .

12
Testing

• A modular approach was followed to test the
  different components of the system
• The sensor node was calibrated with a help of a
  digital thermometer
• The Image Algorithm was tested through MATLAB.
• The GSM link was tested separately.
• The Zigbee modem was interfaced with the RS232
  and wireless communication verified.

13
Costing
E Box 180 ATMega 8 Microcontroller
2 Temperature Sensor 0.6 Humidity
Sensor 10 Vibration Sensor 5 X Bee
module 15 CMOS Camera 7 GSM
Setup 25 Stepper Motor 9 Miscellaneo
us Components (Cables etc.) 15 There are 28
sensor nodes in each eBox zone. Total cost of
one eBox zone Cost at eBox tower 28 Cost
of one node Deployment cost 1192.8
The overhead cost for replacement of faulty
sensor nodes is estimated at 43. Deployment
cost is 100
14
Network Topology

• Addressing scheme - Every node has a unique
  pseudoID which is calculated as follows
• The innermost nodes are assigned consecutive
  natural numbers starting from 1(to a maximum of
  9). Every node in the outer circle will have a
  pseudoID consisting of the parents pseudoID
  appended with a 1 or a 2 to be identified as
  a left or a right child. In this manner, the
  pseudoID of every node is calculated (shown in
  figure) and programmed into the mote.
• Each eBox zone (circle) has 28 nodes.
• Each circle covers a forest area of 1.13 sq.
  km.
• Distance between concentric circles is 200m.

15
Economic Perspective

• Where our project is directed
• The project is directed to be developed into a
  commercial product
• Who will use our project?
• Governments
• Private Companies requiring forest data for all
  kinds of forestry, forest management and forest
  consultancy.
• Research organizations/NGOs for forest
  preservation/conservation.

16
Our Business Plan

• Our objective is to end up as a product as well
  as services-based organization.
• In the initial stages, we only intend to sell the
  system, which is our product.
• Next, we plan to move into the deployment of the
  system and, training and providing the staff
  needed at the central workstation.
• Maintenance and upgradation of the system would
  also be taken care of by us.

17
Business Plan

• At this point, we will also be a data provider
  for organizations needing forest data.
• With the already deployed sensor network forming
  the backbone and each node having multiple and as
  yet unused sensor input channels, we will be able
  to provide customer-specific data.
• Ultimately, we would also want to be able to
  provide forest
• management and forest consultancy thus, making us
  an
• end-to-end solutions provider for new forest
  based enterprises being set up.

18
Microsoft and Forest Watch

• Another useful idea which will go on to make
  Forest Watch a commercial product, is the use of
  MSN Virtual Earth.
• With this, we plan to provide forest maps having
  forest isotherms (temperature) and forest
  isohyets (precipitation).
• These will be useful to specific organizations,
  for all kinds of forestry practices, forest
  management and forest consultancy.

19
An Alternative to GSM - In case a GSM network is
not available Judges Feedback

• Small VSATs (about 0.65m diameter) can be mounted
  on the eBox tower.
• The VSATs are connected to the eBoxes through
  IDUs (In Door Units). The IDUs support IP and are
  connected to the eBoxes through the Ethernet
  port.
• Thus, the eBoxes are connected to the Central
  Workstation via satellite.
• Rates of data transfer being reasonably fast,
  forest images can be easily transferred to the
  Workstation now.

20
Conclusion
Forest Watch is a real-time system, cheaper than
existing ones. Forest Watch cuts down huge
economic and environmental losses. The physical
quantity acquired as data is customizable to
specific needs. All the above show that Forest
Watch and the related services will together
prove to be a profitable product and at the same
time, lead to preserving, protecting and
enhancing the environment. A WIN-WIN SCENARIO!!!
21
Recommendation
We demonstrated our system to the Ministry of
Environment and Forests. They have given us a
green signal to go ahead with the prototype of
the system and have appreciated our efforts in
this direction.
22
Work Distribution

• Manish ( Electronics and Communication)-hardware
  design, testing and procurement
• Vamsi (Instrumentation and Control) and Manav
  (Computer Engineering)-software modules and
  hardware interfacing.
• Pulkit (Instrumentation and Control)-Routing
  protocol, Positioning algorithm and
  microcontroller firmware.

23
References

• 1Jason Lester Hill, System architecture of
  wireless sensor networks,2003 University of
  California, Berkley
• 2Paul Horowitz, Wienfield Hill, The Art of
  Electronics
• 3Microsoft MSDN Documentation
• 4John Murray, Inside Microsoft Windows CE, MS
  Press
• 5www.embedtronics.com
• 6www.mathworks.com
• 7www.avrfreaks.net

24
THANK YOU!