Principal Investigator: Milind Tambe
Other Researchers: Fernando Ordonez
Security at major locations of economic or political importance and transportation infrastructure is a key concern around the world, particularly given the threat of terrorism. The protection of important locations includes tasks such as monitoring all entrances or checking all inbound traffic or patrolling trains and buses. However, limited security resources prevent comprehensive security coverage at all times, which allows adversaries to detect and exploit patterns in selective patrolling or monitoring, e.g., they can plan an attack by avoiding their knowledge of existing patrols. Randomizing schedules for patrolling, checking, or monitoring is thus an important tool in the police arsenal to avoid the vulnerability that comes with predictability. In developing an automated program for randomization, we must address address three key challenges. First, we must provide distinct weights to different actions based on their complex costs and benefits. For example, if an attack on one part of an infrastructure will cause economic damage while an attack on another could potentially cost human lives, we must weigh the two options differently—giving higher weight (probability) to guarding the latter. Second, we must address the uncertainty in information that security forces have about the adversary. For example, while there may be a certain probability that a hard-core, high-capability terrorist group may be planning an attack on infrastructure, there may be a higher chance for a local gang, with lower capability and other motivations, that may be planning an attack. Third, we must take into account adversary’s reaction to our randomized strategy.