How does ultrasonic sensing compare with camera-based occupancy detection?
Ultrasonic sensing is ensuring full privacy of all occupants, technology is detecting the micro movements (sub mm movements). Next generation ultrasonic sensors for interior sensing are capable of detecting human breathing, even when wearing a sweater or in the case of a child under a blanket. Signal processing on-top of ultrasonic sensing is less extensive vs camera image processing, due to this ultrasonic sensing is very well suitable as a low cost and low power alternative for fusion technology to camera for in-cabin monitoring.
Ultrasonic sensing when paired with in-cabin interior microphones may create an advance array of sensors capable of monitoring the car interior for micro movements, temperature changes (ultrasound speed in the air depends on temperature) and interior noise in audible & ultrasonic spectrum (inaudible noise), indicating defective parts requiring repair or maintenance.
Pairing of ultrasonic micromovement detection with driver/occupant audible expressions may improve quality of driver/occupant status monitoring, especially when raw audio data and raw ultrasonic data are processed by AI trained to detect multiple driver and/or occupant health conditions, based upon fusion of audio and ultrasonic data.
Which scenarios are most challenging for rear occupancy detection?
Critical is the coverage of the full rear seat area without blind spots. A single sensor without a blind spot under all possible assembly conditions is not feasible. In the case of next generation car cabins, ultrasonic speakers used as a source of directed audio in combination with interior microphones (capable of both audio + ultrasound signal reception) could provide extensive sensor & actuator array serving multiple purposes – directed audio, voice control, micro movement detection, temperature sensing and acoustic/ultrasonic in-cabin noise monitoring.
Intention is not develop perfect single purpose sensor, while use existing actuator and sensors in novel way providing the max coverage without additional overhead.
How have recent IC improvements enhanced performance?
IC improvements are allowing system integrators to collect true AI data from sensors and benefit from sensor edge processing. Sensor raw data AI processing at Fusion ECU is a way to deliver continuously evolving functionality capable of addressing corner cases as fast as possible (evolving car interior monitoring performance based upon sensor raw data), the only disadvantage is high computational complexity linked with high current consumption.
Edge processing in sensors is a way to reduce overall current consumption related to complex AI processing (sensor raw data processing) running at Fusion ECU. Edge processing in sensors should detect basic patterns that justify waking the Fusion ECU, it is a way to visibly reduce the power consumption, it is essential to deliver max efficiency per battery kWh.
What future applications could emerge from highly sensitive ultrasonic sensing?
There is potential in the combination of ultrasonic speakers (for directed sound delivery, technology capable to create private/isolated sound space for each car occupant) and ultrasonic microphones (microphones capable of operating in freq range ~20Hz up-to ~70kHz) into one multifunctional sensor and actuator array.
Such technology may allow next generation directed audio delivery, cabin micro movement detection, temperature monitoring, voice control, noise cancellation and interior monitoring for various defects and leaks accompanied with increased noise in ultrasonic and/or audio parts of the spectrum.
Interested in exterior vehicle sensing technology?
With a pass to InCabin Europe, you’ll also get full access to our
co-located sister event, AutoSens. Find out more here >>
