The IFAS project set out to create a prototype for interior monitoring of vehicle occupants in high‑ and fully automated cars. The goal was to detect, interpret and classify driver seat positions and activities, especially out‑of‑position situations that pose a safety risk in accidents. To achieve this, the team combined high‑resolution radar and optical sensors in a sensor‑fusion architecture and developed a risk‑assessment model for non‑driving tasks. The final prototype was required to be robust, validated, and ready for integration into safety‑activation modules.

CMORE Automotive GmbH, a rapidly growing German automotive supplier with around 200 employees, led the technical work. The company supplied its proprietary automotive framework, C.FRAME, which synchronises data from multiple sensor sources in real time. C.FRAME had previously handled data rates of up to 1 GB/s from ADAS sensors and supported cross‑sensor calibration, scene management, and KPI reporting. In the IFAS project, C.FRAME was extended to integrate three radar systems: a Silicon Radar unit, a Texas Instruments mmWave radar, and an Acconeer A111‑001‑T&R sensor. Laboratory tests showed that the mmWave radar could reliably record pulse and breathing signals under controlled conditions. Subsequent vehicle tests in a VW Golf and a Mercedes E‑Class confirmed that the radar maintained stable performance while driving, although vibration‑induced inconsistencies in data synchronization were observed and later addressed through software adjustments.

The project was organised into three main work packages: (1) concept and laboratory set‑up, (2) system development and integration, and (3) system testing and evaluation. CMORE handled the sensor‑system build, data collection, and remote integration into a test vehicle, while the Technical University Ingolstadt and the Technical University Chemnitz were contracted for test‑case generation, data acquisition, functional development, and robustness testing. Remote support proved essential during the COVID‑19 pandemic, allowing CMORE to coordinate the final integration and data‑sync optimisation from a distance. The system was successfully installed in a test rig at Ingolstadt, where the fusion of radar and camera data was demonstrated and the prototype’s ability to detect driver posture and activity was validated.

Performance metrics reported include stable detection of pulse and breathing in both lab and vehicle environments, and the ability to handle high data throughput (up to 1 GB/s). While specific numerical thresholds for seat‑position classification accuracy were not disclosed, the project achieved its primary objective of a functional prototype capable of real‑time monitoring and risk assessment. The robustness tests confirmed that the system could operate reliably under varying driving conditions and sensor noise.

Collaboration involved CMORE Automotive as the lead partner, with the Technical University Ingolstadt and the Technical University Chemnitz providing specialised research support. The project was funded under the German federal grant 19A19009C, covering personnel costs of €132,953.39 and external research services of €112,073.11. The consortium worked over a period that included laboratory development, vehicle testing, and remote integration, culminating in a validated prototype ready for further development and potential commercial deployment. The outcome supports CMORE’s broader strategy of offering engineering, data, and platform solutions for autonomous, connected, and electric mobility, positioning the company as a key partner for future mobility innovations.