Artificial Intelligence (AI)-based Accident Prevention and Anomaly Detection System for Light Rail Safety

I&T Solution Artificial Intelligence (AI)-based Accident Prevention and Anomaly Detection System for Light Rail Safety
(REF : S-0710)
Trial Project
Solution Feature
  • Artificial Intelligence-based: learnable thus adaptable
  • Privacy Protection: No facial recognition and no video analytic function, and fully complied with personal data (Privacy) Ordinance
  • The AI can forecast the trajectories of high-risk passengers or anomalous object spatiotemporally
  • Completely stand-alone system in one integrated unit that can be flexibly installed at Light Rail trackside
  • Fast detection in a full 3D space, with necessary information such as wheelchair user or size of human (child / grown-up) provided
Trial Application and Expected Outcome
  • Equipped with AI analytics to differentiate abnormal cases from normal train service scenario, with 4 months deployment period to capture all necessary scenarios for such differentiation
  • On-site site reconnaissance, data collection and integrability study for system design assumptions, planned deployment and AI training
  • Development of alert logic and trigger alarm to Duty Operators for further action, with snapshot and location information records timely disseminated and recorded for immediate / follow-up actions
  • Flexibly movable stand-alone solution with always-on full privacy protection features, customized to Light Rail safety operational guidelines
  • On-site support for full stack AI hardware, software, and system fine-tuning to meet the performance and accuracy requirements
Info on I&T Solution Provider
Solution Provider:AkiraKan Limited
Address:West Wing 2/F, 822 Lai Chi Kok Road, Lai Chi Kok, Cheung Sha Wan, Kowloon, Hong Kong
Contact Person:OOI Ghee Leng
Position:Technical Business Development Manager
Tel: +852 8199 9086
Email: info@akirakan.com
Webpage: www.akirakan.com

For details of the above I&T solution, please contact the I&T solution provider.