Graduation Internship: Reinforcement Learning for Logistic Optimization

Boskalis is currently looking for a student to start their graduation project with us. Logistic planning is a complex task involving numerous dependencies and a chain of events, especially in the large-scale projects that Boskalis conducts. Many projects share common challenges, such as aligning our assets, accounting for external dynamic factors, and meeting the objectives desired by venture partners. Traditional planning methods involve heuristic calculations guided by experience or rules of thumb, which are difficult to manage manually. Multi-Agent Systems (MAS) and Reinforcement Learning (RL) are promising fields worth exploring, as they have shown success in solving NP-hard problems through simulation and verification.

The research project will require the student to conduct a literature review on developing such systems for our defined real-world projects. This will involve finding or building appropriate simulations to run these algorithms. Key design choices need to be made, such as defining the elements of the RL agent(s) and their environment. With a variety of RL algorithms and frameworks available, an educated choice must be made to address our specific problem effectively.

In this research you will

• Work with other data scientists and meet with multi-disciplinary operational engineers to understand the challenge.
• Conduct literature research on MAS and RL and how they can be used for making schedules.
• Formulate the problem in a mathematical model.
• Develop efficient generic methods for making schedules for the chosen methods and demonstrate how they can be scaled up for future needs.
• Report your final findings, demonstrating your design choices and reproducibility of results.

Your qualities

• You are doing a university master’s degree in computer science, AI, mathematics, or other related field.
• Affinity with Python and version controlling.
• Able to develop, train, and evaluate AI models, using TensorFlow, PyTorch, Keras.
• Theoretical understanding of MAS and RL, practical knowledge is not requested but pre.
• Proficiency in writing reports and findings, and presenting results.
• It is preferred to be working at our office in Papendrecht 3 days a week.