Purple Transform recently disclosed receiving funding from Innovate UK for its pioneering work in developing AI models to identify individuals at risk of suicide on railways. The tech company is using the funds for its ongoing project, Project SAIVE (Suicide Avoidance via Intelligent Video Examination), which uses three AI-enabled technologies to evaluate people's behaviour and assess suicide risk.
It alert staff, mental health professionals and police, providing each with advice on how to respond. The period from 2020 to 2021 saw a total of 2,057 suicide attempts and 247 deaths by suicide on the railways. The disruption caused by these instances, combined with the trauma inflicted on the families, friends, rail staff and witnesses, culminates in each incident having an estimated cost of £1.67m. Purple Transform is working in collaboration on Project SAIVE with Govia Thameslink Railway and experts in mental health management and technology from Lancaster University to mitigate this.
The AI being developed utilises psychology, behavioural science and computer vision algorithms. Smart cameras are trained to look for indicators of potential suicide attempts, and risk scores are generated through deep learning which has been trained on pre-existing footage to detect certain behaviours. The system also alerts staff to potential incidents, providing them with instructions on how to proceed and similar alerts can also be sent to mental health professionals and specially trained railway staff from the British Transport Police.
CEO of Purple Transform, Gregory Butler stated, 'We are really proud of our partnership with Lancaster University and Govia Thameslink Railway and pleased it has been recognised by Innovate UK. Every vulnerable individual deserves support, and we are committed to developing ethical AI that has a positive impact on society by creating better human outcomes.' Dr Mahsa Honary from Lancaster University Management School further expanded that by identifying behavioural patterns in highly vulnerable individuals, they can effectively train the AI to intervene efficiently and save lives.
Project SAIVE operates via SiYtE, the proprietary AI platform of Purple Transform which analyses and visualises real-time anonymised data from organisational infrastructures - including CCTV cameras, IoT sensors and third-party feeds - using AI and machine learning techniques.