About this role
Geselecteerde taal: EN - Sustainable - Circular and industrial construction - Go to Mobility and logistics - - Efficient drug development - Medical devices - Work health - - - Accurate geo-location with nanosatellites - PERSEUS wind turbine radar interference assessment tool - Military use of space - Integrated combat capabilities - - Semicon and quantum - - - TNO Traineeship - Talent development programme - Team Polar - Newsroom - Collaboration - Projects on commission - Public-private - TNO Fast Track - TNO Ventures - Innovation centres Geselecteerde taal: EN Locatie: Helmond Werken op afstand: Hybrid Opleidingsniveau: Bachelor Master Uren per week: 32-40 hours/week Vacancy 1285 About this position Within the MARQ Digital Lab, you will develop a context-aware AI assistant for a driving simulator. The assistant will leverage a Large Language Model (LLM) that receives real-time contextual information from the simulation, including the road environment, traffic situation, vehicle status, and driver behavior (e.g., eye tracking and driver inputs such as steering, throttle, and braking). Based on this information, the AI assistant should be able to answer driver questions and proactively provide warnings, explanations, and driving advice. You will design and implement a software architecture that connects the driving simulator (SCANeR) to a locally hosted LLM running on the GPU infrastructure of the MARQ Digital Lab. The project will investigate how simulation context can be efficiently supplied to the LLM, determine the update frequency required for real-time interaction, and evaluate which open-source LLM is best suited for this application. The final result will be a working proof-of-concept of an AI-powered driver assistant that combines natural language interaction with context-aware support during driving simulations. The solution will serve as a foundation for future research into AI-assisted ...