Patrolling has remained as one of the most enduring operations conducted by security entities globally. Law enforcement officers and private security agencies deploy personnel throughout their given jurisdiction in order to protect their stakeholders -- be it infrastructure, property, or the general public. Patrols are deployed to enforce laws, deter crime, and maintain public safety and trust.
There are two challenges for security entities when it comes to deploying their patrols: limited resources and predictability.
For security agencies, having more personnel available for patrols would allow them to be more effective in protecting their stakeholders. This is especially true for entities with large areas of jurisdiction, such as law enforcement. Unfortunately, more personnel would require a higher budget. Entities have a limit in how many patrols they can deploy. Not every area will be guarded at any point in time.
The second challenge is the need to avoid predictability. Perpetrators and other would-be adversaries can observe where the patrols are being deployed, and recognize patterns that they can exploit. If criminals have predicted where the patrols would be deployed at a certain time, they can identify and attack vulnerable areas. Once patrol patterns are identified, their effectiveness drops severely. The challenge for security entities is to deploy their patrols in the most optimal way while trying to avoid creating noticeable patterns.
The University Police Force (UPF) is the official security division of the University of the Philippines Los Baños (UPLB). The UPLB campus is open to the public, and visitors and locals frequently visit the public spaces within campus, mixing in with students and staff. In order to prevent crimes and policy violations, foot and vehicle patrols are deployed. The UPF deploys their patrols in specified zones within campus, and throughout their shift, patrols randomly move from place to place within their zones.
Giving the patrols this much free reign incorporates randomness in their movements. There is a smaller chance of patterns occurring when patrols decide where they go next, and when they go there. Free reign also prevents the patrols from excessive fatigue, as a strict movement schedule would put them under significant pressure and make them lose their sense of free movement.
This approach, however, also has its drawbacks. The lack of explicit route orders make it harder to monitor patrols. It will be difficult for the UPF to guarantee that high priority areas are sufficiently visited, and some areas may be left unguarded for too long without their knowledge. The patrols themselves are also susceptible to their personal biases. Personnel may visit some areas too often, while secluded areas they don't like going to are left vulnerable. Over time patrols would also find it more difficult to actively randomize their movements, increasing their tendency to develop predictable movement patterns.
Randomization is a key solution for avoiding predictability. Having patrol routes and timings that vary regularly would make it harder for adversaries to predict patterns. The uncertainty presented to adversaries is oftentimes effective in discouraging them from attempting an attack. By having no explicit timings and routes, the UPF has incorporated randomness to some extent in their patrol movement. UPF's implementation, however, is done manually, without aid from scheduling systems or randomizer programs.
The lack of assigned routes and letting patrols decide for themselves makes it difficult to assure that the patrols' moves are efficient. The UPF can compose an itinerary for personnel to follow, but the assignments are difficult to randomize by hand. As studies have shown that humans cannot be fully random, the randomization needs aid from a scheduling system. Several agencies, both from the government and private sector, have employed scheduling software to help them make optimal assignments, and the aim of this study is to develop a proposed scheduling system for UPF to use.