To develop AI on Predictive Maintenance and Performance Optimisation for Main Low Voltage Switchboard

I&T Wish To develop AI on Predictive Maintenance and Performance Optimisation for Main Low Voltage Switchboard
(REF: W-0447)
Matched I&T Solution
Trial Project
Summary and Challenges
  • With more smart power analyzers and sensor technologies installed in electrical distribution systems inside government buildings, more electrical data become readily available continuously through online monitoring system such as Power, Quality & Energy Monitoring Systems (PQEMS) for data visualization.
  • This project is to make use of online monitoring system and  real-time sensor data of various government buildings to build AI model for predicting potential faults and abnormalities, optimizing performances, analyzing load distribution and thus enhancing overall building operational stability and energy optimization.
Expected Outcome
  • This project involves
  1. building database and AI detection algorithm based on historical and real-time data, available knowledge database and professional research,
  2. training and refining machine learning model, and
  3. implementing the AI model and initiating operation analysis and alert when abnormalities found.
  • The AI model shall be able to study including but not limited to busbar high impedance detection, early fault finding & potential damage to load equipment, energy loss estimation and energy distribution analysis.
  • The AI model shall provide decision supporting system that can
(a) formulate and recommend fault prediction and energy optimisation,
(b) analyze results and trigger alarm when abnormalities found and
(c) generate reports for record and recommendation purpose.
Expected Trial Duration 10-month
Contact Information
I&T Wish Proposer:Electrical and Mechanical Services Department (EMSD)
Contact Person:HON Chun Keung, Steve
Position:Building Services Engineer/Electrical & Mechanical and Building Services Technology Development/4
Tel:2808 3263
Email: ckhon@emsd.gov.hk
Upload Date 2023-03-29
Closing Date 2023-04-24