Now Reading
AI & Game Theory Can Now Predict Where Poachers Are Likely To Strike Next


AI & Game Theory Can Now Predict Where Poachers Are Likely To Strike Next


A simple theory put forward by famous mathematician John Nash, known as Game Theory, has many applications today, than they were when it was newly proposed. According to the theory, any game has an equilibrium a collection of strategies, one for each player, such that no player can win more by unilaterally switching to a different strategy. It attempts to predict how an individual within a group will choose between different strategies, when the outcome of the situation depends on how everyone else in the group behaves. The difficulty is that you can’t work out an individual’s optimal strategy without knowing what all the others will do.



This theory has many application in complex systems of daily lives into the factors that govern the chance and decision-making. Interestingly, the same theory is being applied to curb poaching of wildlife animals.

India And Wildlife Poaching

According to a study by the Wildlife Protection Society of India (WPSI), India lost 260 leopards in the first six months of 2018 of which 90 were killed by poachers. Anish Andheria, president, wildlife conservation trust and member of Maharashtra state wildlife board said, “The figure of 90 confirmed leopard poaching cases is probably a gross underestimate. Electrocution and poisoning are quiet killers. These deaths generally remain undetected.”

The Government of India enacted the Wildlife Protection Act 1972 with the objective to effectively protect the wildlife of this country and to control poaching, smuggling and illegal trade in wildlife and its derivatives. The Act was amended in January 2003 and punishment and penalty for offences under the Act have been made more stringent. But even then, there have been animal poaching crimes in the country, and the criminals are not subject to punishment.

How Game Theory Can Solve The Problem Of Poaching

Models based on game theory can be used to know about the locations and frequencies of animal poaching. Through this, they can assist conservation practitioners in devising more effective patrol strategies. This can be done by simulating the dynamic interactions and adaptations between patrols and poachers while considering the movement patterns of the wildlife species of interest.

University of Southern California has already developed an artificial intelligence software called Protection Assistant for Wildlife Security (PAWS). The software uses game theory and mathematical modeling techniques to help increases chances of saving wildlife.

The data that PAWS uses is of past patrolling and poaching activities. Using this data, it predicts the locations and routes of the possible patrols in the future. For predicting the poachers’ behaviour, they use game-theoretic reasoning and route planning. Based on the poachers’ behaviour model, the software calculates a randomised patrolling strategy, in the form of a set of patrol routes and the probabilities of taking each route.

PAWS then suggests patrol routes sampled from this strategy to the patrollers. The input data includes information like contour lines that describe elevation, terrain information such as lakes or drainage, base camp locations, and monitor previous patrol observations. Based on the input data, current animal population distribution can be estimated. Because wildlife patrols and poaching attacks happen frequently, the activity can be modelled as a repeated game. Once the model has been built, the poachers’ behaviour can be modelled using the wildlife data and then optimal patrol strategy according to the game model, behavior model, and additional patrolling constraints can be estimated.

In game theory, strategic decisions are made based on the behaviour of the opponent. In combating poaching as well, there is a strategic interaction between the poachers and the conservation agencies. In this game, there are two types of players with conflicting

Interests, and can be called as: defender and poacher. Each player wants to take actions to satisfy their intentions.

See Also

Poachers would want to capture wildlife without having their snares confiscated by patrollers, and their opponent, patrollers, would want to find snares before they capture any animals. Each player needs to reason about his or her opponent's potential actions and play intelligently, similar to game theory.

For example, if the patroller always takes the same patrol route every day, the poachers will always be able to avoid the patroller and will always succeed in placing snares in the unpatrolled locations. Thus, it is in the defender's best interest to play unpredictably. Instead of randomly choosing which patrol route to take, the defender should choose patrol routes that visit more important locations more often. Which patrol routes to consider and how to randomly choose among these patrol routes is called the defender strategy or the patrol strategy. Ideally, for the defender, poachers would be deterred from locations with high animal density since they are often patrolled, and the poachers would be reluctant to place snares in areas with low animal density since the chance of successfully capturing an animal in those areas is low.

Outlook

Game theory can take inputs from various of these data and predict where and when the outbreak of an animal poaching can be. Authorities can then not just catch the offender, but also prevent the act of animal poaching. It will greatly help us to preserve animal preservation.

Paul Oxton, founder of Wild Life Wild Heart Foundation (WHWF) had rightly said, “The future of wildlife and the habitat that they depend on is being destroyed. It is time to make nature and all the beauty living within it our priority.”



Register for our upcoming events:


Enjoyed this story? Join our Telegram group. And be part of an engaging community.

Provide your comments below

comments

What's Your Reaction?
Excited
0
Happy
0
In Love
0
Not Sure
0
Silly
0
Scroll To Top