The day after tomorrow: optimising the autonomous vehicle system

10 May 2017

Dr Albert Lam Yun-sang (Croucher Fellowship 2009) is a research assistant professor of Electrical and Electronic Engineering at The University of Hong Kong (HKU).

Lam grew up in Hong Kong and was educated in the city from kindergarten through to PhD but it was a gift from his parents that first triggered his interest in physical science.

“When I was ten years old I was given my first PC- a large desktop and it made me interested in logic and mathematics,” he says and his bachelor’s degree at HKU was in Information Engineering which he chose because of his interest in communications and networking. It was to remain the foundation of his later research.

When he started his PhD (at HKU), it was his intention to stay in the area of information engineering but his supervisor, Professor Victor O K Li, encouraged him to study something more general and to “think outside the box.” So Lam looked at the field of optimisation, the mathematical techniques commonly used for solving complex problems in the field of engineering.

“We are always looking for the optimal state in any system,” explains Lam, citing a simple example based on his more recent research. An autonomous vehicle system with thousands of vehicles requires a system to devise an optimum route for each vehicle, given the start and end point, road conditions congestion status and other traffic requirements.

Conventional theoretical mathematics is very powerful if the problem fits into certain structures. However not all problems fit into these structures, which is where CRO and other metaheuristics are important.

Lam became interested in metaheuristics for optimisation, which was inspired by nature. Many heuristic algorithms are very specific and problem-dependent, while a metaheuristic is a high-level problem-independent algorithmic framework that provides a set of guidelines or strategies. An example of nature-inspired metaheuristics is genetic algorithms and ant colony optimisation which attempts to mimic how an ant colony migrates from one place to another to search for food. It is used for solving computational problems which can be reduced to finding good paths through graphs.

“People had developed metaheuristics inspired from biology and physics but no-one had tried to mimic chemical processes,” says Lam who achieved a major breakthrough during his PhD by co-inventing Chemical Reaction Optimisation (CRO) a metaheuristic for optimisation, inspired by the nature of chemical reactions. The natural process of transforming unstable molecules with excessive energy to stable ones through a sequence of elementary reactions is embedded in CRO to solve optimisation problems and can be applied to tackle problems in both the discrete and continuous domains.

"Conventional theoretical mathematics is very powerful if the problem fits into certain structures. However not all problems fit into these structures, which is where CRO and other metaheuristics are important," says Lam.

This research was different from the traditional concept of formulating a problem and looking to solve it. Instead, it devised a method that could be applied to many complex problems, which is consistent with his own way of undertaking research using creative ideas.

While Lam’s PhD was focused on methods, his post doc work at University of California, Berkeley was more focused on green energy and smart grids, a hot topic in 2010.

“I achieved some useful research ideas in new solutions for obtaining optimal power flow,” says Lam modestly.

Optimal power flow is a classic power system problem and it’s highly complicated because it involves not only matching generation with demand in all conditions but also the design of how the power flows through the network to minimize power loss.

“We identified certain structures in part of the network where conventional optimisation techniques and algorithms could be applied effectively,” says Lam. He extended this work when he returned to Hong Kong in 2012 and took up a position at Hong Kong Baptist University. He combined his research into the smart grid with the concept of the smart city, integrating the smart grid with transport.

“The autonomous electric vehicle can become an active and smart user of the grid,” says Lam explaining that in a vehicle energy network (VEN), electric vehicles are used to transport energy to remote locations where infrastructure is inadequate by using a dynamic wireless charging network in the road network.

This area dominates his current research at HKU where he took up his current positon in 2015 and it has enormous potential for transforming the efficiency of transportation in cities. “If you think of Uber and the future of smart city- people will just make an Uber-type request on their mobile phone and the system would assess requests and allocate an autonomous vehicle and facilitate ride-sharing,” says Lam.

Lam has designed a system for public transport using autonomous vehicles and believes it combines the advantages of private taxis and mass public transport and it’s not just a theoretical model.

“It’s a very practical system if only the government would one day allow autonomous vehicles,” he says and while he has not received much interest from authorities in Hong Kong, in Singapore, they are already testing autonomous vehicles and the Mainland sees it as a potential tool for easing traffic congestion.

“I think that Mainland China and Singapore are possibly more aggressive in examining technology solutions to traffic congestion while dynamic wireless charging is now possible and is being tested in South Korea,” he says.

Lam says in the future, all vehicles could be wirelessly charged dynamically with renewable energy and individuals would have no need for their own vehicles- they could just call one on their mobile phone and he admits it can be frustrating working in Hong Kong where there is a lack of an open mind for new technologies.“It would be better if the government would support the ideas or at least open the door and look at them,” he says.


Dr Albert Lam Yun-sang (Croucher fellowship 2009) is a research assistant professor of Electrical and Electronic Engineering at The University of Hong Kong (HKU). He was a postdoctoral scholar at the Department of Electrical Engineering and Computer Sciences of the University of California, Berkeley, California, USA. He received both his BEng degree in Information Engineering and his PhD degree from HKU in 2005 and 2010, respectively. He is an active member of the IEEE Computational Intelligence Society. In January 2017, he was appointed as the Chapters Subcommittee Chair and in December 2015 Lam served as the Mobile App Chair at the 2016 IEEE World Congress on Computational Intelligence (WCCI2016). In 2012, he returned to Hong Kong to work at Hong Kong Baptist University and accepted his current role at HKU in 2015. His current research interests include optimisation and evolutionary computation as applied to smart grids and smart cities. More specifically Lam’s recent work has been focused on optimal power flows, autonomous vehicles, vehicle energy networks and intelligent transportation systems. 


To view Dr Lam's Croucher profile, please click here