IoT Thrust Seminar | Human-in-the-loop learning for Smart Cities
IoT and machine learning are transforming urban life, particularly by optimizing smart transportation and service systems to reduce delays and enhance efficiency. However, since humans serve as both sensors and users (e.g., choosing routes or service providers), traditional systems often struggle with inefficiencies. This talk introduces human-in-the-loop learning solutions that integrate machine learning with human-aware mechanism design, improving both system performance and user experience.
The first part of this talk analyzes crowd-sourced navigation platforms such as Google Maps and Waze. While these systems effectively aggregate real-time user data to recommend fastest routes, they frequently fail to adequately explore time-dependent traffic conditions on alternative pathways. This exploration-exploitation imbalance can create systemic inefficiencies as too many users converge on the same 'optimal' routes. Our solution introduces strategic information disclosure mechanisms that intelligently guide user routing decisions while maintaining provable performance guarantees - specifically, ensuring travel cost never exceeds twice the optimal solution.
The second part explores IoT-enabled service facilities, such as restaurants and hospitals, which periodically sample and update online queue length information to help customers decide when to visit. Surprisingly, providing more frequent—or even real-time—updates does not necessarily benefit either the system or its users. Leveraging this insight, we develop optimal algorithms to determine the ideal frequency of information updates, significantly improving system performance and user experience.
Lingjie Duan is the Associate Head (Research) and a Tenured Associate Professor at the Engineering Systems and Design (ESD) Pillar/School of the Singapore University of Technology and Design (SUTD), which was established in collaboration with MIT. He received his Ph.D. from The Chinese University of Hong Kong in 2012 and his Bachelor's degree from Harbin Institute of Technology in 2008. In 2011, he was a visiting scholar at the University of California, Berkeley.
Dr. Duan’s research focuses on resource allocation in IoT networks, distributed machine learning, multi-agent systems, and cybersecurity management. He has published 170 papers, including six ESI Highly Cited Papers, in leading venues across engineering, AI, and business disciplines. He has secured 17 research grants from government and industry, totaling CNY 88 million. Since 2023, he has co-led the SUTD Wireless Innovation Centre, managing CNY 22 million in industry funding for edge AI and IoT research.
Dr. Duan has received numerous prestigious accolades, including recognition among Stanford University’s Top 2% Scientists since 2021 and as a 2022 Highly Ranked Scholar by ScholarGPS. His awards include the First Prize Paper Award from the Computer Academy of Guangdong, the SUTD Excellence in Research Award, the 10th Asia-Pacific Outstanding Young Researcher Award from the IEEE Communications Society, and a Finalist for the Hong Kong Young Scientist Award.
He currently serves as an Editor for IEEE/ACM Transactions on Networking and as an Associate Editor for IEEE Transactions on Mobile Computing. Previously, he was an Editor for IEEE Transactions on Wireless Communications and IEEE Communications Surveys and Tutorials, as well as a Guest Editor for the IEEE Journal on Selected Areas in Communications. Dr. Duan was the General Chair of the WiOpt 2023 Conference and has served as a Technical Program Committee (TPC) member for top-tier networking conferences, including IEEE INFOCOM, SECON, and ACM MobiHoc.