Test Engineer – Interior Detection

Test Engineer – Interior Detection
The key function for automotive interior sensing solutions is to detect small children left behind in parked and locked cars under hot climate conditions. Such a system can alert the driver through vehicle light/sound alerts or smartphone app notifications that something is wrong, avoiding heat stroke injuries or in worth case, death. Such systems can not only detect human presence in the cabin but also the seat position of the human and secondarily also theft attempts, achieving a 3-in-1 system. Today’s solutions are mainly based on mm radar wave solutions but in the future there will be sensor fusion concepts, combining radar sensors with cabin cameras. This will enable an interior detection functional growth, for instance to support body position detection, human size classification, child seat detection etc.
To establish such systems, massive data collection from real vehicle environment is needed. Both as a basis for designing human body movement detection algorithms and also to validate the performance of these algorithms and the alignment against OEM requirements.
The test engineer is responsible for maintaining interior detection test methods and for executing the tests described. The test variety is multi-dimensional as a function of occupant age (new born, 1 year, 3 year, 6 year, 10 year, 5%-ile female, 50%-ile male etc.), seat position (up to 7 different seats), body position/orientation, body direction (forward facing vs. backward facing), child seat type/size, vehicle state (parked, in motion) and vehicle variant (5-seater, 6-seater etc.). Data collection tools are used to gather interior sensor data, combined with cabin video stream and metadata. Such tools are developed in-house and under continuous improvements. Therefore, the test engineer also needs to have programming competence (C and Python) to maintain and update such tools to support new test cases and future carline cabin sensor configurations.
The test engineer also needs to interact with our interior detection algorithm engineers to analyse test data and classify data to achieve good coverage for both true positive and false positive use cases (for instance “moving objects” classified as humans). The data analysis is made by in-house developed tools to achieve a semi-automated tool chain. Also here, programming skills (typically Python) is a must. In the final stage of the development, the test engineer mainly focus on collecting “a validation data set” from real car interior testing with humans to verify and document the system performance and to agree a “start of production sign-off” with customer.
Driver license is required since some tests are conducted with the vehicle on the road. The test engineer also needs to travel to the premises of the OEM to get access to test vehicles. Since test methods and test results often needs to be discussed and negotiated with our customer (OEM), a positive, pro-active win-win attitude will be an important factor for long term success.