Crime and Incident Watch for Smart Cities: A Sensor-Based Approach

Beyond the many advantages which Smart City brings, the issue of security and privacy remain very important concerns. As things get more interconnected, cyber orchestrated crimes such as cyberterrorism may become more prevalent. For example, an autonomous vehicle or drone may be used to commit acts of terrorism, such as driving or flying into a crowd. Responses to such incidents need to be swift and likewise smart. Law enforcement needs to be equipped with the necessary tools to respond speedily and even automatically, equally taking advantage of the Internet of Things (IoT). We propose a Sensor-Based Crime and Incident Watch for Smart Cities (SBCI-Watch), leveraging Human Activity Recognition (HAR). SBCI-Watch may be used to automatically detect and possibly report occurrences of public disturbances (likely caused by acts of terrorism or similar crimes) to foster swift response. Slightly similar reports exist, but they are focused on different topics, such as helping law enforcement to pick suitable officers to respond to crime scenes based on proximity, but without attempting to automatically detect crimes. We are using the sensor-based approach; it is less privacy intrusive compared to the vision-based method, which is dominant in the public surveillance area but also more privacy intrusive and more expensive. To the best of our knowledge, our work is the first to address the problem in this fashion, and the results illustrate the viability of the SBCI-Watch.