Towards a SoIS model for University Surveillance

University surveillance systems that use sensors (cameras, GPS) operate in dynamic and heterogeneous environments. Interaction with other systems is, therefore, an important aspect to consider. We should consider these types of systems as a system of information systems (SoIS). Modeling this type of system and integrating its components presents a significant challenge, as various factors may hinder achieving the desired security solution. It is fundamental to consider the interoperability property of a SoIS as a requirement for implementing a university surveillance system. Interoperability achievement is an obstacle when heterogeneity between systems is present turning achieving a SoIS-based surveillance solution into a critical issue to address. Our proposal, an architecture designed to interoperate with various systems related to university security, is a crucial step in addressing the complex interoperability issues. We developed an architectural model and implemented a mediator to analyze the level of interoperability that can be achieved. We then applied this mediator to a risk scenario that could escalate into a high-impact incident for the university community, demonstrating its practical viability. We identified the systems involved in university surveillance and how their heterogeneous data can be managed to ensure the necessary communication for appropriate risk detection and prevention of high-impact incidents. In summary, through exploratory study, we validate our solution. We achieved syntactic interoperability through the video stream protocol. We achieved semantic interoperability through communication between the camera and license plate recognition system with 96.79% accuracy. We also achieved semantic and pragmatic interoperability between camera and suspicious behavior detection systems with better adaptation to the actual context, with a precision of 95. 24% and a recall of 76. 92%. IT can prevent high-impact situations by analyzing suspicious behavior scenarios.