AiBLE Lab 2 PhD Positions (# of pos: 4)

TU Delft is a top tier university and is exceedingly active in the field of Artificial intelligence. The AiBLE lab will investigate how AI can be developed and used in complex real-world scenarios to help reach effective, transparent, and lasting design decisions and agreements. In AiBLE we also incorporate human feedback in the loop to iteratively improve decision-making and drive behaviour changes.

AiBLE will focus on two important challenges in the built environment: energy transition and circularity. By using AI to augment human intelligence and support negotiation continuously while adapting to and driving human behaviours, our lab will close the policy-practice gap responsibly and inclusively. We aim also for the acceleration of adoption and acceptance of responsible and inclusive AI.

AiBLE has 4 PhD positions (for position 2 and 3 we have already selected new colleagues), we currently are actively recruiting for position 1 and 4:

Position 1) How can the use of AI in multi-actor settings contribute to a collaborative circular built environment strategy? This project will first analyze and then model multi-actor preferences with regards to the circular transition in the built environment. This will then inform the design and experimentation of AI technologies in multi-actor real-world settings to examine how design decisions can be made collaboratively to reach more circular outcomes. In this iterative process, it is crucial to combine existing and newly-obtained data from multiple sources with feedback from users to allow continuous improvement.” (Daily supervisor: Dr. Tong Wang, Dr. Luciano Siebert; Promotor: Prof. dr. Paul Chan; Place: Management in the Built Environment, Faculty of Architecture and the Built Environment)

Position 4) Reinforcement learning for human-AI interaction in the built environment. How can AI be used for learning to influence built environment users to make more sustainable choices? This project will explore the potential of AI-based agents such as digital assistants and social robots that are embodied and embedded in the built environment to form an ongoing relationship with users and influence their daily decisions to make more sustainable choices for circularity and energy transition. Techniques from interactive machine learning and reinforcement learning will be developed and applied to the problem of long-term interaction to create agents that can maintain a trustworthy and persuasive relationship with users. (Daily supervisor: Dr. Frank Broz, Dr. Luciano Siebert, Dr. Tong Wang; Promotor: Prof. dr. Catholijn Jonker; Place: Interactive Intelligence Group, Faculty of Electrical Engineering, Mathematics and Computer Science)