At TU/e, we solicit applications for PhD students in an industry-driven project. With our industry partner, we investigate more realistic, but also easily usable planning tools for mid-sized transport companies. At the core of all algorithms and planning tools stands the possible increase of sustainability, e.g., by reducing empty mileage and unused capacity.
Project descriptionWith this project, we aim to develop planning solutions using both traditional methods from operations research and artificial intelligence. The developed tools will have the potential to make road transport both more efficient, more resilient, and more sustainable, and will also address driver shortages, ongoing electrification and automation of road transport, and collaboration and competition in the transportation market.
During an early phase of the project, the students will jointly get to know the industry partner and the challenges of such medium-sized road transport companies. Further, the students will become acquainted with the daily job of planners, and will perform some preliminary analyses. We will also use this time to break down the project into multiple smaller research projects.
Some exemplary research projects include:
- Learning the planners' decision-making process: We will use historic data on decisions (using inverse optimization) and/or experiments to elicit information about the planning process, such as informal soft constraints but also considered future events.
- Improving the interaction between planners and tools: In their day-to-day work, planners face a multitude of messages coming in (e.g., existing inefficiencies, new orders, or delays of loads requiring replanning) with different urgency and impact. Using experiments, we will investigate how different communication measures increase or decrease the planners efficiency.
- Developing next-day planning algorithms: Transport companies must coordinate not only their trucks, but also their loads and their drivers. All three are subject to a multitude of regulations (e.g., country-specific vehicle permits, driver rest time). We will develop algorithms for next-day planning using efficient algorithms such as branch-price-and-cut as well as heuristics.
- Developing dynamic planning algorithms: We will combine classic OR techniques with reinforcement learning methods to allow real-time planning of trucks, drivers, and loads.
TU/e and the teamThe PhD students will be supervised by dr. Albert H. Schrotenboer, dr. Layla Martin, and prof. dr. Tom Van Woensel. Albert and Layla are both assistant professors with a clear focus on planning and operating transport and logistics, and Tom is a full professor for freight logistics.
Our industry partner is an international logistics service provider employing more than 400 people headquartered in Utrecht (around one hour by car from TU/e, costs covered). They offer transport, flexible storage and logistics optimization, and act as a transporter (2PL), freight forwarder (3PL) and handle the complete management of logistics transport activities (4PL). Our industry partner strives to offer the best possible sustainable transport solutions based on honesty and respect for people and society.
The project, the supervisors, and eventually also the PhD students are embedded in TU/e's Operations, Planning, Accounting, and Control (OPAC) group, the Eindhoven AI systems institute (EAISI) of TU/e, and the European Supply Chain Forum (ESCF). OPAC uses methods from operations research and operations management on a wide variety of problems, and currently hosts around 50 PhD students from various backgrounds. EAISI brings together all AI-related activities throughout TU/e and will support the PhD students by providing an enriching environment. ESCF works on fostering collaborations between industry and academia, among others in their program lines 'ESCF Circular' and 'AI planner for the future'.
Target profileWe are looking for PhD students with a keen interest making transport more sustainable. Students shall have prior experience in optimization, operations research, decision support, and/or artificial intelligence, for example proven by a suitable master degree (e.g., Industrial Engineering, Econometrics, Applied Mathematics, Computer Science). Ideally, applicants have already finished their master degree or are expected to finish early 2023 to allow a project start before summer 2023.
Candidates must have experience in programming, using languages such as C++ or Java. Ideally, candidates have experience with mathematical modeling and implementation in CPLEX or Gurobi, implementing efficient heuristics, and/or machine learning.
Due to our industry collaboration, we particularly look for outgoing, communicative students with willingness to learn the Dutch language. In particular during the first phase of the project, the students will spend several days per week at our industry partner in Utrecht.