Two PhD positions in Artificial Intelligence for Data-Driven Logistics

Updated: 2 months ago
Deadline: 08 Aug 2021

Are you interested in adapting techniques from Artificial Intelligence (specifically Deep Reinforcement Learning (DRL)) to support people who plan transportation and logistics operations in practice? Do you want to help develop cutting-edge techniques, i.e. DRL algorithms inspired by the latest breakthoughs in the field and/or hyperparameter tuning and algorithm selection for DRL using AutoML? Are you motivated to make these techniques applicable for our partners from practice, including ASML and Vanderlande? We are looking for two PhD students in Operations Research and Artificial Intelligence with a focus on those topics.

Current breakthroughs in Artificial Intelligence are exciting for games like Go and Chess, where it is crucial to anticipate unknown moves of the opponent. When making logistics decisions, it is equally important to anticipate the arrival of new data (e.g., orders, delays, and disruptions).
For many such problems, Deep Reinforcement Learningalgorithms like AlphaZero have been demonstrated to be game-changers. The logistics sector recognizes the opportunities and is eager to adopt deep reinforcement learning. However, companies struggle to translate the abstract possibilities of deep reinforcement learning into solutions for their own logistics problems.

To address this challenge, we aim to develop a toolbox that contains the tools to rapidly model and solve logistics problems with Deep Reinforcement Learning, preferably using zero-code solutions. The toolbox will be tested by using it to solve concrete logistics problems at one or more of our ten project partners.

You, as a successful applicant, will perform research on the project outlined above in an international research team. You will discuss possible your ideas with partners from practice.
You will report research findings at international conferences and workshops, and in high-quality scientific journals.  The research will be concluded with a PhD thesis. A small teaching load (on average about 10%) is part of the job description.

We are looking for two PhD students, one will be oriented more towards integrating Deep Reinforcement Learning techniques into an easy-to-use toolkit. This position will be based in
the Operations Planning Accounting and Control (OPAC) Group and will be supervised by Willem van Jaarsveld and Remco Dijkman. The other position will be oriented more towards the development of the Deep Reinforcement Learning techniques themselves. This position will be based in the Information Systems (IS) Group and will be supervised by Yingqian Zhang and Willem van Jaarsveld.

Both positions will be based in the DynaPlex project and require strong collaboration with each other, with another PhD position at the University of Twente, and with our partners from practice, including ASML and Vanderlande.

View or Apply

Similar Positions