Ultra-Wideband In-Cabin Sensing

Use Cases – UWB Radar

UWB Radar – In-Cabin Sensing

In addition to UWB ranging for measuring signal propagation time between two UWB nodes, UWB technology can also be used as a Pulse Doppler Radar. In this scenario, a UWB node sends out a UWB pulse and receives its reflections from the surrounding space. The result is a complex pattern of overlaying reflections from various reflectors. The Doppler effect allows for the resolution of the smallest movements of objects.

Due to its high resolution, UWB can detect minute movements of living beings, including the movement of the chest during respiration (see figure) or the movement of the heart muscle. However, measuring the movement of the heartbeat is only feasible under laboratory conditions. In practical settings, the movement of the heartbeat is often overshadowed by various other movements, such as breathing, and consequently, it gets lost in the sensor signal.

The ability to detect these movements leads to numerous use cases.

UWB Sensor Set-up

Embedded AI is a leader in developing algorithms for detecting vital parameters in the automotive environment using UWB Radar. For in-cabin monitoring, there are various approaches to the placement of UWB anchors. External UWB nodes:

  • left in the image – established variant
  • right in the image (object of research) – current research: external UWB nodes moved to the side doors to support both the exterior and interior monitoring.

In both variants, two additional UWB nodes are placed in the roof lining. Other placements and a different number of nodes are also possible depending on the use cases.

The detection algorithms work in a distributed manner. Some algorithms are calculated on the UWB node itself, including signal processing algorithms and feature extraction, and in some cases, even AI algorithms for event detection in the signal. Allowing some use-cases without additional processing hardware. Certain features are sent to the control unit (ECU). The ECU merges the intermediate results of the individual nodes and can thus execute sensor fusion algorithms allowing better spatial detection and accuracy.

Use Case: Vital Sign Detection, Breathing

Embedded AI has multiple approaches for in-cabin sensing to determine vital parameters from respiration monitoring scenarios and is a leader in research and development in this field.

Use Case: CPD – Child Presence Detection

The Problem: Heatstroke

Especially during summer time or on days with high solar radiation, temperatures can become life-threatening in a locked car, particularly for babies and animals. The critical temperature of 42°C inside the car is reached after only 30 minutes at an outside temperature of 26°C and after 10 minutes at an outside temperature of 35°C. According to current statistics (and estimates), about 800 children worldwide die each year from heatstroke because they are left in a closed car on a hot day (underreported for animals).

The issue is so relevant that the European society for vehicle safety, EURO-NCAP, introduced both interior monitoring and driver monitoring as criteria for assessing passenger safety in 2023. In the USA, these technologies are even more highly valued. A bill called the “Hot Car Act” was passed by Congress in 2021.

The Solution: UWB Radar

UWB technology can detect vital signs of humans and animals and locate them inside the vehicle. UWB Radar has a distinct advantage over optical methods and even conventional FMCW Radar. This is because UWB Radar signals, characterized by longer wavelengths, can effortlessly penetrate materials such as clothing and blankets. Consequently, UWB technology excels at sensing the movement of living beings. Due to the lower frequencies compared to FMCW Radar, UWB has lower requirements for HF components, thus keeping material costs low.

Vital parameters are detected in the UWB Radar signal through AI-based recognition algorithms that run resource-efficiently in real-time on the microcontroller of the UWB sensor electronics.

Child Presence Detection (CPD) is undoubtedly the main motivator for investing in the UWB system in cars. The following sections highlight additional functionalities that can be achieved with the established UWB technology.

Use Case: In-Cabin Sensing during Driving

Driver/Occupant Monitoring

The system can also ensure that vital parameters of occupants and animals are monitored during the drive. The driver is alerted if vital parameters of the occupants reach dangerous levels.

Emergency Services Alert

If dangerous values in vital parameters occur for the driver, emergency services can be alerted automatically.

Vital Sign Information in Case of an Accident

In the event of an accident, data can be automatically transmitted to emergency services to save valuable time in life-saving efforts (who was sitting where and until when they were breathing, abnormalities in vital parameters of individual persons, e.g., irregular heartbeat, shallow, fast, or absent breathing).

Passive Driver Vital Monitoring

For autonomous driving up to Level 3, a person with a driver’s license must control the car or intervene in an emergency. During long, monotonous drives, fatigue can quickly set in for the passive observer – a latent danger. For this level of autonomous driving, the UWB system provides support in monitoring the (passive) driver by detecting fatigue, sleeping, atypical situations (such as a heart attack), or respiratory arrest.

Use Case: Intrusion Detection

Additional cost savings for the vehicle’s system costs arise by using the UWB system when in-cabin monitoring is also conducted in a parked and locked vehicle. The UWB system, through movement detection, can reliably detect break-ins. The costs for conventional alarm system sensors can thus be eliminated.

Unlike in-cabin monitoring during the drive, the “alarm system” mode operates with a significantly lower frequency of measurement cycles to minimize standby power consumption.

Interested?

We hope we could provide you with a glimpse. If you're interested in seeing demos or live applications, feel free to get in touch with us.

  • email: po[at]embedded.ai
  • phone: +49 351 6533 3808