Parallel Workshops
Cases of self-harm and violent acts within correctional facilities are an ongoing concern. Conventional means of monitoring inmates present significant problems related to 1) privacy concerns; 2) operations that are labour-intensive; and 3) high setup and maintenance costs. A novel intelligent system has been developed to enable autonomous real-time monitoring of inmates with the capability of automated detection of anomalous events of interest to the user. These anomalous events include self-harm behaviours, violent acts, and hazardous changes in life signs. The driving design motivation for the system is to enable continuous monitoring of these anomalous events while respecting inmate privacy and personal data rights. In this presentation, the system design is introduced and results from in-house testing are discussed. Plans for deployment in real-world correctional facilities are also presented.