Project SAIVE
Lead Participant:
PURPLE TRANSFORMATION GROUP LIMITED
Abstract
Project SAIVE's (Suicide Avoidance via Intelligent Video Examination) objective is to boost rail industry productivity and simultaneously deliver social value by addressing the prevalent issue of suicide prevention, with a particular focus on the context of railway. Beyond the direct impact (due to delays, penalties, cost of replacement services to deal with disruption), the nature of the incident causes enormous productivity and social costs in terms of staff (driver and station) and onlookers' trauma.??
The approach to addressing the challenge of suicide intention detection in the context of rail will rely on innovative, three levels of AI-powered capabilities implemented with Purple Transform's platform SiYtE. SiYtE is a suite that integrates data from cameras and sensors, analyses, visualizes and feeds back insights concerning events and incidents, patterns and metrics of interest. The first layer incorporates expert knowledge from psychology and behavioural science into computer vision algorithms implemented on smart cameras to look out for indicators of potential suicide attempts. The second layer complements this approach by using deep learning models trained on existing footage to provide prediction and risk scoring via picking up ineffable, non-linear patterns in behaviour. The third layer of AI serves to direct the alerts to the right staff with suggestions for actions and potential communications with suitably trained personnel to assist with the incident, whilst also preventing incident exacerbation by alerting train drivers, British Transport Police and mental health professionals. Through such preventive actions, the SAIVE is expected to boost productivity of rail operators.?
Project SAIVE is a collaborative partnership led by Purple Transform (expertise in computer vision, data integration and visualisation, integration with workflow, benefit modelling) joined by University of Lancaster (expertise in mental health management using technology), working with Govia Thameslink Railway (rail operator that knows and understands these challenges).
The approach to addressing the challenge of suicide intention detection in the context of rail will rely on innovative, three levels of AI-powered capabilities implemented with Purple Transform's platform SiYtE. SiYtE is a suite that integrates data from cameras and sensors, analyses, visualizes and feeds back insights concerning events and incidents, patterns and metrics of interest. The first layer incorporates expert knowledge from psychology and behavioural science into computer vision algorithms implemented on smart cameras to look out for indicators of potential suicide attempts. The second layer complements this approach by using deep learning models trained on existing footage to provide prediction and risk scoring via picking up ineffable, non-linear patterns in behaviour. The third layer of AI serves to direct the alerts to the right staff with suggestions for actions and potential communications with suitably trained personnel to assist with the incident, whilst also preventing incident exacerbation by alerting train drivers, British Transport Police and mental health professionals. Through such preventive actions, the SAIVE is expected to boost productivity of rail operators.?
Project SAIVE is a collaborative partnership led by Purple Transform (expertise in computer vision, data integration and visualisation, integration with workflow, benefit modelling) joined by University of Lancaster (expertise in mental health management using technology), working with Govia Thameslink Railway (rail operator that knows and understands these challenges).
Lead Participant | Project Cost | Grant Offer |
---|---|---|
  | ||
Participant |
||
PURPLE TRANSFORMATION GROUP LIMITED |
People |
ORCID iD |
William Wood (Project Manager) |