AI to identify potential suicide attempts via Montreal's subway surveillance system

The AI system, which is still in the development phase, leverages approximately 2,000 surveillance cameras installed at subway stations. The Société de transport de Montréal (STM - Montreal's public transit company) and the Center for Suicide Intervention (CRISE) are collaboratively developing the system. The AI examines the video feed from the cameras in real-time and identifies individuals who may be experiencing extreme stress.

Professor Brian Mishara, the director of CRISE and a professor of psychology at the University of Quebec in Montreal (UQAM), who is a globally recognized expert in suicide prevention, explained the workings of this system to CBC, a public television network. "We have clues, but obviously, humans can't monitor hundreds of screens all day to identify such behaviors," said Mishara. Thus, AI is being employed, which relies not on facial recognition algorithms but on characteristic behaviors.

Professor Wassim Bouachir of the University of TÉLUQ, a Quebec-based university specializing in distance learning, described some of these behaviors to the "La Presse" newspaper. "People who spend a lot of time staring into the tunnel, individuals approaching the edge of the platform, those regularly crossing the yellow line, and those who let several trains pass by - these are all indications that may suggest a suicide attempt," explained Bouachir.

Mishara has been conducting research on suicides in the Montreal subway for many years. In his 1999 paper published in the Canadian Journal of Psychiatry, he scrutinized coroner reports from 1986 to 1996 related to 129 suicides in the Montreal subway. He noticed that the majority of individuals who committed suicide had mental health issues, and 81 percent of them had expressed their intent.

In 2016, Mishara published an analysis of the behaviors of individuals who committed suicide in the Montreal subway, caught on camera from 2010 to 2013, in the scientific journal "BioMed Central Public Health". This study was the first of its kind, as reported in a UQAM press release. Mishara noticed that individuals who left their belongings on the platform and repeatedly paced between the edge and the wall could be potential suicide victims. "Based on these two behaviors, captured through the cameras, almost a quarter (24 percent) of the suicides could have been prevented," Mishara claimed in the UQAM press statement.

The ongoing pilot program is considering additional factors. The aim is to construct the AI algorithm to potentially alert station security in time to intervene, or even inform the subway conductor to stop the train before entering the station. Bouachir mentioned to "La Presse" that, following the completion of accuracy tests, the system should be ready for implementation within a maximum of two years.