I-2503 – PHD IN EXPLAINABLE AI FOR DATA-DRIVEN PHYSIOLOGICAL AND BEHAVIORAL MODELLING OF CAR DRIVERS

Temporary contract | 14+22+12 months | Full-time/40h | Belval Are you passionate about research? So are we! Come and join us

How will you contribute?

The evaluation by test drivers of new tire designs in real situations is a key part of the tire development process. In this context the role of the tester and the feedback it gives about the car and tires behaviour is determinant for the tire engineers to build perfect tires. This feedback however is very subjective and varies from one driver to another, which makes it hard to build proper models without understanding the influencing human factors.

Generally, the evaluation of physiological and behavioural data is a key part of understanding human-centric systems. In this context, the role of subjective and objective measures, and their integration, is crucial for creating robust and interpretable AI models. Today, the complexity of multi-dimensional data and the need for transparency pose significant challenges, which this PhD position aims to address via the very concrete case of simulated driving.

You will:

• Explore methods to disentangle and integrate objective and subjective data sources to separate overlapping signal components and isolate meaningful patterns in complex datasets.
• Develop AI and XAI (explainable AI) models to analyse and interpret physiological and behavioural data, in combination with vehicle telemetry, to predict a driver’s subjective tire handling rating. Knowledge-driven models using ontologies will be explored here with XAI models.
• Investigate the precision and completeness of physiological data, proposing innovative and advanced solutions to handle missing data, outliers, and/or inconsistencies.
• Incorporate subjective measures, such as ratings and feelings, into AI frameworks.
• Contribute to the project’s broader goals by integrating insights into actionable outcomes.

Activities:

• Conduct extensive background literature analysis, focusing on disentanglement, AI, XAI, and human-centric data analysis.
• Elaborate validation use-cases and participate in planning data collection studies under controlled conditions.
• Automate processes and explore state-of-the-art solutions for filtering, noise reduction, and kernel-based techniques.
• Present your findings at academic conferences and publish research in peer-reviewed journals.
• Write a PhD thesis in the field of AI and human-centric data analysis.
• Take part in research and training activities as part of your PhD program.
• Participate in outreach activities of LIST

The Luxembourg Institute of Science and Technology (LIST) is a Research and Technology Organization (RTO) active in the fields of materials, environment, and IT. By transforming scientific knowledge into technologies, smart data, and tools, LIST empowers citizens in their choices, public authorities in their decisions, and businesses in their strategies.

In 2024, LIST and GOODYEAR signed the second phase of their partnership, 2024-2029, building on the outcomes of their previous collaboration and entering new technological areas as well. GOODYEAR-LIST Partnership 2.0 embraces Luxembourg's National priorities, such as Sustainability, Digital transformation and Circular economy, through the execution of six Strategic Research Programs: Data Science, Tire as a Sensor, End-of-Life Tire Valorization, Sustainable Materials for Non-Pneumatic Tires, Sustainable Materials for Next Generation of Pneumatic Tires, Structure-Process-Properties Relationships.

As part of our Data Science strategic research program, we are looking for a PhD candidate in artificial intelligence and data analytics to analyse vehicle telemetry data along with physiological and behavioural data of test drivers and build explainable predictive models out of it. The research team spams Luxembourg, Europe and the USA and will make use of world-leading car simulator facilities. Work in the project comprises human-factors research, artificial intelligence and data analytics.

Do you want to know more about LIST? Check our website: https://www.list.lu/

Is your profile described below? Are you our future colleague? Apply now!

Education

• A Master’s degree or Engineer diploma in Computer Science, Artificial Intelligence, Data Science, Machine Learning, or a related field.

Experience and skills

• Strong knowledge of AI, Machine Learning, data-science (e.g., neural networks, deep learning, autoencoders, GANs, active learning, etc.);
• Knowledge of explainable AI and Knowledge Graphs with ontology (e.g., RDFS, OWL);
• Demonstrated experience with common advanced signal processing techniques (e.g. denoising autoencoders, robust PCA, wavelet transforms).
• Experience in methods to disentangle and integrate data sources (e.g., InfoGAN, β-TCVAE, TopDis / Topological Disentanglement, Independent Component Analysis (ICA), Variational Autoencoders (VAEs), and matrix factorization techniques);
• Experience with programming in Python and machine learning frameworks (e.g., TensorFlow, PyTorch, Scikit-learn);
• Interest in or experience with XAI methodologies and their applications;
• Excellent communication skills and the ability to work collaboratively in an interdisciplinary environment.

Language skills

• Proficiency in English (written and spoken). Knowledge of French is an asset but not mandatory.