A new collision avoidance system (CAS) that is inspired by the way insects avoid collisions has been developed by researchers from the American Chemical Society.
The new system is much simpler and more energy-efficient than its predecessors, making it a more effective and safer option for use in vehicles.
Current systems often rely on complicated signal processors to make sense of images captured in poor lighting conditions, or use radar or LiDAR sensors, which are difficult to miniaturize and require a lot of power. However, insects can easily avoid collisions with each other without using such technology, even at night.
The researchers designed an algorithm based on the neural circuitry used by insects to avoid obstacles. Instead of processing an entire image, the system only processes one variable: the intensity of a car’s headlights. This means that the detection and processing units can be combined, making the overall detector smaller and more energy-efficient.
The sensor used in the new system is made up of eight photosensitive “memtransistors” constructed from a layer of molybdenum disulfide, which is organized onto a circuit. The sensor is only 40 µm2 in size and uses just a few hundred picojoules of energy, which is tens of thousands of times less than existing systems.
In tests, the detector was able to sense a potential two-car accident two to three seconds before it happened, giving drivers enough time to take critical corrective action.
The researchers believe that this new detector could help improve existing CASs and make them safer.