Data-Driven Control Strategy for Automotive Cabin Air Quality Management

Ensuring clean and healthy cabin air is vital for passenger well-being and comfort. Traditionally, modern vehicles use air quality sensors to detect air pollution and automatically activate air recirculation. While effective, these sensors come with drawbacks, including added complexity in integration and wiring, delayed response times, and gradual performance loss due to contamination and wear.

This work introduces a software-based approach for intelligent air quality management, eliminating the need for dedicated air quality sensors. By utilizing data already available within the vehicle – such as driving dynamics, navigation information, and other relevant signals – this method enables predictive control of the air recirculation flap. Extensive on-road measurements are conducted in urban, rural, and highway settings, capturing particulate matter both inside and outside the vehicle, as well as gas concentrations outside the vehicle. The results reveal that stationary air quality data fail to accurately represent on-vehicle conditions. Moreover, distinct driving situations, traffic influences, and environmental factors are identified as major contributors to exterior air quality deterioration.

The proposed data-driven strategy proactively limits pollutant ingress by anticipating upcoming air quality scenarios and adjusting the flap position before contaminants enter the cabin. Its reliance on existing vehicle data makes it cost-effective and highly scalable for series production, supporting rapid deployment across different vehicle platforms without additional hardware. This approach represents a practical step toward sensor-free cabin air quality management with tangible advantages for vehicle manufacturers and passengers, directly aligning with the focus of the conference session "Vehicle Innovation, Safety and Sustainability".