Professor Sebastian Stein

Research groups

Research interests

• Citizen-Centric Artificial Intelligence Systems
• Mechanism Design and Incentive Engineering in Multi-Agent Systems
• Applications of AI in smart energy, transportation, electric vehicle charging and disaster response
• Reinforcement Learning
• Human-AI Partnerships

Current research

Sebastian works in the area of artificial intelligence and in particular multi-agent systems. He is interested in highly dynamic, heterogeneous systems where multiple self-interested actors (including human users and intelligent software agents) come together, interact and possibly pursue conflicting objectives. To deal with these challenging settings, Sebastian's research focuses on a range of techniques:

• Mechanism Design and Incentive Engineering: This looks allocation and payment mechanisms that incentivise desirable behaviours despite the presence of self-interested agents.
• Sequential Decision Making with Uncertainty: This considers dynamic settings with potentially incomplete information and high uncertainty, where an autonomous agent needs to choose actions sequentially to maximise some objective.

Sebastian is passionate about applying these techniques to a range of important application areas and key societal challenges, including:

• Smart Mobility: Artificial intelligence holds promise to revolutionise the way we travel. Sebastian is interested in applying incentive engineering to design efficient smart mobility systems that allow citizens to complete journeys seamlessly and on-demand, using more sustainable forms of transport rather than the currently prevalent model of widespread vehicle ownership.
• Electric Vehicle Charging: Electric vehicles, when coupled with renewable electricity generation, are key to reducing carbon emissions from transportation. However, widespread use will likely cause considerable strains on the electricity distribution networks. Sebastian has applied mechanism design to this setting to enable smart charging schemes that allocate electricity efficiently within constrained settings.
• Crowdsourcing: Increasingly, people and intelligent algorithms work together to solve complex problems, ranging from coordinated crowdsensing activities during disasters to large citizen science efforts. Sebastian is using machine learning and sequential decision-making approaches to deal with the uncertain and noisy data provided by human contributors.
• Cloud Computing: Running computationally intensive analytics tasks is challenging in settings where resources are constrained and uncertain, e.g., during disaster response operations, in IoT networks or in edge clouds. Sebastian is interested in applying techniques from game theory to model how multiple self-interested actors can share these resources and how AI techniques can be used to deal proactively with uncertainty.

Research projects