Internship | Interpretable Design-Space Exploration

Internship | Interpretable Design-Space Exploration

Many industries, including automotive, healthcare, industrial automation, aerospace, and high-tech manufacturing, rely on complex and distributed Cyber-Physical Systems (dCPS) to deliver innovation and well-being to society. These systems integrate vast networks of software and hardware components, operating across heterogeneous platforms. As these systems grow in complexity, companies find it increasingly challenging to evaluate and understand the impact of their design decisions in terms of key system qualities, such as cost and performance. To successfully design the next generation of innovative systems, new, efficient, and scalable design-space exploration methodologies for industrial dCPS are needed.

What will be your role?

TNO-ESI is working together with the Parallel Computing Systems (PCS) Group and the high-tech equipment industry to research new scalable and efficient model-based design-space exploration methodologies that automatically search the large design space of possible systems configurations to find the configurations that satisfy requirements and provide the best trade-offs between qualities like cost and performance.

Together with Thales, TNO-ESI has developed a methodology for synthesis of component compositions and deployment in the context of microservice architectures, a design space exploration problem with a huge design space. The developed methodology automatically searches the design space and presents the best configuration(s). However, it does not provide any insights about what makes these configurations good. This is a missed opportunity to learn about what really works in the system and reduces trust in the proposed solutions. The key idea in this project is to look at similarities and differences between the design points with high in fitness and distill learnings about the system. This can be used to enrich the framework to not only provide the best configurations, but also a rationale describing what makes those configurations good, which helps build trust in the solution.

Possible research questions are:

• What are the most effective techniques for extracting interpretable insights from DSE results in dCPS?
• How can these insights be validated for correctness and usefulness in industrial settings?

Tentative steps involved in this could be:

• Investigate the state-of-the-art in distilling insights into systems and software in a data-driven way. Classify existing solutions and identify research gaps.
• Extract design points and fitness their fitness scores from the existing system synthesis methodology.
• Develop an approach to extracting insights from the best data points. This may be done using conventional analysis methods or leveraging recent advances in large language models.
• Evaluate the developed approach for a number of handcrafted cases with known properties to establish their correctness and accuracy, then use an industry case study to evaluate scalability and applicability.
• Write and defend a thesis and optionally publish results as an academic paper.

What we expect from you

You are a Computer Science, Software Engineering, or Computer Engineering MSc student looking for a graduation assignment. You are excited about investigating how to extract system insights from data, leveraging traditional analysis techniques or AI technologies, such as large language models. Experience with software development, Linux, Python, large language models, and optimization (e.g., using meta-heuristics) is an advantage.

What you'll get in return

You want an internship opportunity on the precursor of your career; an internship gives you an opportunity to take a good look at your prospective future employer. TNO goes a step further. It’s not just looking that interests us; you and your knowledge are essential to our innovation. That’s why we attach a great deal of value to your personal and professional development. You will, of course, be properly supervised during your work placement and be given the scope for you to get the best out of yourself. Furthermore, we provide:

• A highly professional, innovative internship environment, within a team of top experts.
• A suitable internship allowance (615 euro for wo-, hbo- and mbo-students, for a full-time internship).
• Possibility of eight hours of free leave per internship month (for a full-time internship).
• Use of a laptop.
• An allowance for travel expenses in case you don’t receive an OV-card.

TNO as an employer

At TNO, we innovate for a healthier, safer and more sustainable life. And for a strong economy. Since 1932, we have been making knowledge and technology available for the common good. We find each other in wonder and ingenuity. We are driven to push boundaries. There is all the space and support for your talent and ambition. You work with people who will challenge you: who inspire you and want to learn from you. Our state-of-the-art facilities are there to realize your vision. What you do at TNO matters: impact makes the difference. Because with every innovation you contribute to tomorrow’s world.

Contact

E-mailadres: david.worrell@tno.nl