Enter the search phrase
Domain: Interior sensing
Service: Methodology preparation, dataset collection and analysis, testing and validation
Automotive OEM needed to select and test driver monitoring solutions for its passenger vehicles. The Robotec.ai team was given the task of testing the solutions and collecting reference dataset for measures related to driver distraction detection (head and eye position tracking). Additionally, we had to test the DMS camera system in three different lighting conditions: sunset directional light, light day conditions with the zenith sun angle and moonlight.
Together with our partners, we prepared an artificial solar system for simulation of multiple lighting conditions. The system allowed us to set up lighting conditions in the laboratory to account for natural light spectral distribution according to the international standards. Our focus was on wavelength characteristics for driver monitoring systems (near infrared).
We adapted the high-end motion capture system consisting of 28 cameras to track drivers’ head and eyes positions. Our DMS tracking system provides the highest resolution motion capture camera on the market with submillimeter level of accuracy.
Our team, together with our manufacturing partners, prepared a detailed and accurate vehicle mock-up, based on CAD files, using a combination of CNC and 3D printing technologies. We prepared a fixation targets scheme and synchronised the light targets component with the reference motion capture system. For this we used our proprietary double synchronisation method providing millisecond synchronisation accuracy.
With human participants (drivers) we accounted for diversity in ethnicity, multiple age groups and occluding features (masks, different types of glasses etc.).
We co-developed ROS 2 gem for Open 3D Engine to enable performant and scalable simulation of agriculture, mining and warehouse robots. Tailored for ROS, it gives full access to the ROS 2 ecosystem.Read more
Dedicated driving simulator dataset collection campaign for generating ground truth datasets drowsiness detection. Our team applied a combination of eye closure metrics, driving performance and physiological data analysis (ECG and EEG).Read more