Graduation Internship: AI & Machine Learning

Boskalis is inviting master's students to join as a Graduation Internship in AI and Machine Learning. The role is onsite in Papendrecht, NL, and it’s a full-time placement within a global leader in dredging and marine infrastructure. You’ll join a vibrant, international team of experts and have a chance to apply advanced theory to real-world data and operations.

The project centers on the current machine-learning-based recommendation system that guides dredging crews in setting up the process in real time. There’s room to push beyond the present approach by exploring physics-informed machine learning, time-series modelling, and adaptive control to capture relationships the existing model may miss. You’ll review relevant literature, make design choices about combining physical knowledge with data-driven models, and plan validation using historical data to guide improvements.

Shaping smarter real-time controls

During the internship you’ll collaborate with fellow data scientists and engage with engineers from different disciplines to understand the challenge from multiple angles. You’ll search the literature on physics-informed ML, time-series modelling, and adaptive control to see how these methods could be applied to this problem. You’ll then formulate a problem definition that blends physical knowledge with data-driven approaches and set up approaches to validate any improvements using historical data. In the end, you’ll develop and demonstrate methods that improve the model’s recommendations and show how they generalize across vessels and operating conditions, concluding with a report that documents your design choices and the reproducibility of results.

Must-haves for the project

• You are pursuing a university master’s degree in computer science, artificial intelligence, mathematics, physics, or a related field
• You have a knack for Python and experience with version control
• You can develop, train, and evaluate machine learning models
• You bring a theoretical grounding in physics-informed ML, time-series modelling, or adaptive control; practical knowledge is not required, but is a plus
• You can write clear reports and present your findings
• Preference is given to applicants who can work from the Papendrecht office for at least three days a week

Nice-to-haves that help your candidacy

• Some hands-on experience with applying physics-informed ML, time-series modelling, or adaptive control to real problems
• Experience conducting literature reviews and turning insights into design choices
• Comfort working in a cross-disciplinary environment with engineers and other specialists

Pay, perks and practical details

The internship includes an internship allowance and a welcoming, challenging work environment. Boskalis offers a dynamic setting where you’ll be part of a diverse team of experts and learn from leadership in the field. If you’re under 36, you can join Young Boskalis, which features monthly social and sporting activities, including events like pub quizzes, yoga, bootcamps, and an annual sailboat race, plus opportunities for networking and knowledge sharing.

Resources you’ll be able to use include supervision from AI and data science experts in the AI department, access to data and compute resources (Databricks) for training and evaluating models, and a low barrier to planning meetings to better understand the problem. You’ll work with established development facilities such as Python, source code control, and packaging, and you’ll have access to other ongoing AI department projects for reference and inspiration.

What you can expect is structured guidance for your graduation internship, a warm welcome to help you settle in quickly, and an internship allowance with plenty of challenges to keep you engaged. The environment is designed to be dynamic, enabling you to learn a lot by being part of a leading company and a diverse team of experts.

Tips to stand out in your application

At the top of your resume, lead with your master’s field and any relevant coursework or projects in Python and version control so the panel sees you’re ready to work with data-driven methods from day one.

Be prepared to describe any hands-on experience you have with physics-informed ML, time-series modelling, or adaptive control, and show how you would combine physical knowledge with data-driven models. If you have conducted literature reviews, share how you translated those insights into concrete design choices for models or workflows.

In interviews, outline how you would validate improvements using historical data and what metrics you would use to assess generalization across vessels and operating conditions. Have a plan ready for how you would structure a small, reproducible study that demonstrates your proposed approach.

Ask a targeted question about how the team plans to generalize the model across different vessels and conditions, and what data sources or tools (such as Databricks) will be used to support validation and deployment.

Joining Boskalis: your path to this internship

Interested candidates should apply by filling in their details and uploading a cover letter and CV on Boskalis’s careers site. If you have questions, you can reach Tim Gouweleeuw, Corporate Recruiter, at [email protected].