Engineering Professor and PhD Students Gained Grand Prize and Gold Awards in Global AI Challenge for Building E&M Facilities – AI Competition

13/10/2022

Driving AI Development in Building Services Industry

Engineering Professor and PhD Students Gained Grand Prize and Gold Awards in Global AI Challenge for Building E&M Facilities – AI Competition

Prof. Wang Zhe (second left) and PhD student Guo Mingyue (center) received three awards at the ceremony on Sept 23, 2022. Also pictured are Prof. Sun Dong, Secretary for Innovation, Technology and Industry (second right), Mr. Eric Pang, Director of Electrical and Mechanical Services (first right), and Mr. Alex Lee of Tencent Cloud (first left). [Download Photo]

An inter-university team led by a School of Engineering (SENG) faculty member and a PhD student team received Gold Awards in the academic group of the Global AI Challenge for Building E&M Facilities – AI Competition, an international event that attracted 120 teams from academia and industry.

The two teams were among the 10 Gold Award teams in the academic group of the competition. The inter-departmental PhD student team comprised Muhammad Zeshan AKBER, Muhammad WAQAR (both from Civil and Environmental Engineering), Murad ABDULLAH, and Syed Awais Wahab SHAH (both from Electronic and Computer Engineering).

The other team include Prof. WANG Zhe, who joined the Department of Civil and Environmental Engineering in October 2021, his PhD student Guo Mingyue (who was a master’s student at Tongji University before joining HKUST in Fall 2022), a former research assistant at the department, as well as a collaborator in the industry. On top of the Gold Award, Prof. Wang’s team received Huawei Most Innovative Use of Data Award, which is one of the five Grand Prizes selected from the 10 Gold Award teams, and the Best Use of Tencent Cloud Award, a special prize for contestants who use and describe usage of the Tencent Cloud Platform (TI-ONE). As a big winner in the competition reaping three awards, the team won a prize money of HK$200,000 altogether. The team was ranked first among all the teams in the academic group, with their solution achieving the lowest root-mean-square-error (RMSE) (see the leader board). Lower RMSE means higher prediction model accuracy.

Six other teams involving HKUST or SENG members also received one Silver Award and two Bronze Awards in the academic group as well as two Silver Awards and one Bronze Award in the open group.

In the competition, participants were required to develop a semantic AI model to predict the cooling demand of a commercial building and shortlisted teams needed to submit a pre-recorded presentation video to present their work. The presentations are evaluated based on semantic AI application, innovation, model design, and resources being used, among others.

“This AI challenge is about building load prediction, which has profound application in building energy system design, operation, and grid management. Because of the significance of this work, lots of studies have been conducted on this topic from researchers worldwide. However, testing on an open dataset and competing with scholars from other institutions – that this competition offers – is the best way to demonstrate the superiority of a machine learning algorithm, than just publishing a paper and claiming the best performance on a private dataset,” said Prof. Wang, who would like to thank his team members for the contribution as well as the School and the Department for the support.

The PhD student team were delighted to receive a Gold Award and enjoyed the recognition that came with it. “We participated in this Global AI Challenge with the motivation to use our AI skills in solving an industrial problem and to contribute toward green industrial solutions by improving energy efficiency of the buildings. The workshop provided us an opportunity to meet experts from technical industries; and discussion with them further motivated us to work hard in solving the posed problem. We were lucky in organizing a well-coordinated team, in which each member turned out to be an AI enthusiast, possessed interdisciplinary skills, and efficiently played his role. We would like to thank this platform that connected us and since then, via discussions, we have been continuously improving our AI knowledge,” said Muhammad Zeshan Akber on behalf of his team.

Jointly held by the Electrical and Mechanical Services Department of HKSAR Government and Guangdong Provincial Association for Science and Technology, the Global AI Challenge for Building E&M Facilities is a global event highlighting AI development and applications in the building services industry. The event is the first and largest AI event related to building electrical and mechanical (E&M) services in the world, with the aim of promoting international innovation and technology ideas through exchange and cooperation. Running since September 2021, the event features a series of activities including an international technical conference, the AI Competition, a pre-contest workshop, and a prize presentation ceremony, which was held on September 23, 2022 at Cyberport.