PhD in Deep Reinforcement Learning for planning of lifespan-extension measures

Updated: over 1 year ago
Job Type: Temporary
Deadline: 17 Oct 2022

Are you curious about making inner-city bridges and quay walls future-proof while keeping their cultural-historical value? Do you want to develop cutting-edge techniques from Deep Reinforcement Learning (DRL) for optimized planning of lifespan-extension measures for bridge and quay wall structures? Are you motivated to make these techniques applicable to specific use cases? We are looking for a PhD student with a focus on those topics.

Many old Dutch cities are characterized by their historic quay walls and bridges. In recent years however several walls have collapsed, and many more are in poor condition. The old structures have deteriorated due to deferred maintenance and changed social, technological, and environmental conditions. Maintaining cultural heritage and, at the same time, preparing them for the future is a huge challenge. Lifespan-extension measures are a promising way to cope with this challenge instead of complete replacement.

We are seeking a PhD candidate for the research project 'Sustainable Circular Life Extension Strategies for Inner-City Bridges and Quay Walls' (STABILITY). This exciting research is part of the Urbiquay Program of the Dutch National Research Agenda (NWA). It is funded by the Dutch Research Council (NWO) and in collaboration with University of Twente, Delft University of Technology, Saxion University of Applied Sciences, the Municipalities of Amsterdam, Den Hague, and Zwolle, and several contractors and engineering firms.

Deep Reinforcement Learning algorithms have demonstrated to be game-changers for complex and evolving problem settings having unstructured, diverse input data and uncertain system states. You, as a successful applicant, will develop an adaptive planning method for lifespan-extension measures based on deep reinforcement learning to account for the dynamics and uncertainties related to further deterioration of structures and the developments in the urban context. Your method will allow municipalities to consider the relevant social, technical, and environmental factors of the complex and uncertain urban environment for deciding on the type, moment, and location of the measures to be applied. The proposed approach should be able to execute multiple scenarios, thus rendering the optimal strategy for the renovation and replacement. With this, municipalities will be able to move from a reactive to a more proactive management approach.



Similar Positions