Assignment type: Graduation
Start date: As soon as possible
Assignment Duration: 6-9 months
Educational Level: WO
Desired Study: Data Science
Can raw data be converted in valuable insights about the asset health, to drive and optimize maintenance? What can state-of-the-art anomaly detection algorithms tell us about the health of our systems?
Vanderlande systems produce a lot of data, which can be a goldmine for data scientists to transform into insights. Focus of this internship is ADAPTO, a complex asset with many sub-components that need to perfectly work together.
ADAPTO produces a lot of log data, a very small part of it is being used already by the Predictive Maintenance team to create valuable insights on the asset health. The team makes large use of LSTM models to achieve this goal, and results so far are very promising.
However, recently new techniques based on the heavily hyped Generative AI are being used for Predictive Maintenance. More specifically, also in the field of anomaly detection, transformer networks and attention learning are being researched and applied.
The goal of this assignment is to take the existing data and techniques (Recurrent Neural Networks) used by the team and investigate whether the use of transformers is a faster and more scalable way to approach predictive maintenance product development.
As intern you will work in the Digital Service Platform department, responsible for driving Vanderlande digital transformation through the creation of Digital Services. They will be working in the Predictive Maintenance team, a multidisciplinary, multicultural and distributed team focused on converting data into insights to drive and optimize Maintenance. The team has Data science, data engineer, software development and frontend development capabilities.
As student you will have ample support and access to the existing knowledge within the team. If the internship is successful, their work will be used by the team to improve and enlarge the existing product.
The student will be trained on the ADAPTO functioning, and will be provided with data to be working with.
The student is expected to:
- Interact with SMEs to fine-tune their ADAPTO understanding.
- Learn how to use the latest transformer architectures and apply them to analyze the data provided to him.
- Validate the model outcome (and possibly improve the model) with SMEs and end-users.
- Draw conclusions on the feasibility and scalability of such an approach, compared to existing method.
A successful internship will deliver a report and a small proof of concept for Predictive Maintenance.
- Data science and ML knowledge
- Python knowledge
- (Preferred) Familiarity with Azure and Databricks
Do you recognize yourself in this challenging profile? And are you looking for an internship/graduation assignment in an organization that has been elected as “Best Employer” for years in a row? Please fill out the application form and upload your resume and cover letter. For more information, contact us by e-mail: email@example.com or Stef Alferink (Campus Recruiter) by phone: +31 413 755 087.