green security
play

Green Security How Can AI Help in Protecting Forests, Fish and - PowerPoint PPT Presentation

Green Security How Can AI Help in Protecting Forests, Fish and Wildlife MILIND TAMBE Helen N. & Emmett H. Jones Professor in Engineering University of Southern California WHAT MIGHT WE LOSE? Murchison Falls National Park, Uganda 2/23


  1. Green Security How Can AI Help in Protecting Forests, Fish and Wildlife MILIND TAMBE Helen N. & Emmett H. Jones Professor in Engineering University of Southern California

  2. WHAT MIGHT WE LOSE? Murchison Falls National Park, Uganda 2/23

  3. WHAT MIGHT WE LOSE? Murchison Falls National Park, Uganda 3/23

  4. PAWS: PROTECTION ASSISTANT for WILDLIFE SECURITY Massive forests (1000 sq miles) to protect, limited security resources : • How to Efficiently Patrol/Protect forests with limited resources • PAWS patrols : Exploit past poaching data, avoid predictability Patrol boat in Bangladesh � Patrol with Rangers, Indonesia � at Global Tiger Conference, 2014 � Trip with WWF, 2015 � 4/23

  5. AI-based DECISION AIDS TO ASSIST IN SECURITY Game Theory Airports 2007 5/23

  6. AI-based DECISION AIDS TO ASSIST IN SECURITY 6/23

  7. AI-based DECISION AIDS TO ASSIST IN SECURITY Game Theory Airports Player B Player B � Paper � Rock � Scissors � 0, 0 � 1,-1 � Paper � -1,1 � 2007 -1, 1 0, 0 1, -1 Rock � Player A Player A � 1, -1 -1, 1 0, 0 Scissors � 7/23

  8. AI-based DECISION AIDS TO ASSIST IN SECURITY Game Theory Airports Canine patrol at 2007 LAX � ( ARMOR ) � 8/23

  9. AI-based DECISION AIDS TO ASSIST IN SECURITY Game Theory Airports Air Marshals Ports 2007 2009 2011 9/23

  10. PROTECT: FERRY PROTECTION DEPLOYED [2013-] 10/23

  11. AI-based DECISION AIDS TO ASSIST IN SECURITY Game Theory Airports Air Marshals Ports Trains 2007 2009 2011 2013 11/23

  12. GLOBAL PRESENCE OF SECURITY USING GAME THEORY 12/23

  13. SOME RESULTS OF GAME THEORY for SECURITY Game Theory in the Field 20 Ticketless Travelers Caught Game Theory 15 Previous Method 10 5 • Game theory vs Previous Method 0 #Captures per #Warnings per #Violations per 30 min 30 min 30 min Miscellaneous Arrests at LAX checkpoints 100 Drugs 80 Firearm Violations 60 40 20 0 13/23

  14. GAME THEORY FOR PATROLS [2013] Congressional Subcommittee Hearing 14/23

  15. PAWS: APPLYING AI FOR PROTECTING WILDLIFE Game Theory + Poacher Behavior Prediction Predicting Poaching from Past Crime Data Learn from crime data Game Theory calculate Poachers randomized attack targets patrols Patrollers execute patrols 15/23

  16. POACHER BEHAVIOR PREDICTION Queen Elizabeth National Park, Uganda 12 years of patrols, 125000 observations Area habitat Ranger patrol frequency How likely is an attack on Animal density Area slope a grid Square Distance to … rivers / roads 16/23

  17. PAWS INITIAL SYSTEM [2016] Game Theory + Poacher Behavior Prediction 12 years of patrols, 125000 observations Dry Season (June-August 2008) 17/23

  18. PAWS PATROLS IN THE FIELD [2016] Trials in Uganda and Malaysia Important Lesson: Geography! Uganda Andrew Lemieux Malaysia Panthera 18/23

  19. PAWS: PROTECTION ASSISTANT FOR WILDLIFE SECURITY [2016] Game Theory + Poacher Behavior Prediction + Forest Street Map 19/23

  20. PAWS: PRELIMINARY EVALUATION Previous Patrol � PAWS Patrol � 0.86 � 0.57 � Human Activity Sign/km � 20/23

  21. PAWS COLLABORATIONS 21/23

  22. AI DECISION AIDS for PROTECTING FORESTS, FISHERIES, RIVERS Protecting Forests, Fish, Rivers FOREST PROTECTION FISHERY PROTECTION RIVER POLLUTION PREVENTION 22/23

  23. AI and GAME THEORY WORLDWIDE FOR SOCIAL GOOD Thank you to sponsors: 23/23

  24. THANK YOU tambe@usc.edu http://teamcore.usc.edu/security 24/64

  25. EVALUATING DEPLOYED SECURITY SYSTEMS NOT EASY How Well Optimized Use of Limited Security Resources? Security Games superior vs Human Schedulers/”simple random” Field Evaluation: Field Evaluation: Lab Evaluation Patrol quality Tests against Unpredictable? Cover? adversaries Compare real schedule Simulated adversary “Mock attackers” Human subject Capture rates of Scheduling competition adversaries real adversaries Expert evaluation 25/64

  26. FIELD EVALUATION OF SCHEDULE QUALITY Improved Patrol Unpredictability & Coverage for Less Effort Patrols Before PROTECT: Boston Patrols After PROTECT: Boston Base Patrol Area Count Count Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7 Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7 PROTECT (Coast Guard) : 350% increase in defender expected utility 26/64

  27. EXPERT EVALUATION Example from ARMOR, IRIS AND PROTECT June 2013: Meritorious Team July 2011: Operational Excellence Commendation from Commandant Award (US Coast Guard, Boston) (US Coast Guard) September 2011: Certificate of February 2009: Commendations Appreciation (Federal Air Marshals) LAX Police (City of Los Angeles) 27/64

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend


More recommend