Range-Finding Rear-Collision Accident Prediction and Warning For Motorcycles with Integrated Signal Vest

Design - Theory

Designing the sensor required balancing the tradeoffs between different sensor types. The available sensor fields are sound, light, and camera. Sound sensors were eliminated immediately due to the fact that this system is meant to be active while at high speeds. The wind noise at highway speeds can disrupt the sound waves sent out and thus negatively affect the accuracy of the sensor. The access sound created by vehicles at high speeds would also have some frequencies that match the frequencies of the sensor and thus produced false return times. Another issue to consider is the fact that at highway speeds, the distance traveled by the car during the time of flight of the emitted sound wave would have to be taken into consideration. A flight time while the car is stopped would be a different distance than the same amount of flight time on the highway if sound was being used. Automotive industry only uses sound-based sensors for low speed systems like parking assistance and back up sensors. For these reasons, sound-based sensors will not be used. Camera-based sensors also fell off the table once the magnitude of the complexity to process received imagery and perform image processing for accident prediction was discovered. Since the system is only supposed to detect if a vehicle will hit the host vehicle, it does not make sense to over produce a sensor by making it a camera-based system. Implementing a system involving a camera and copious amounts of image processing to simply determine the distance to a vehicle would unnecessary. Eliminating camera-based sensors from the table leaves only light-based sensors. In the realm of light-based sensors in automotive, there are two main types of sensors. The first is LiDAR, which is used for mapping a given area. This method was rejected for the project for the same reasons as the camera, too much unnecessary processing for the task at hand. This sensor system would perform the task needed without needing large amounts of processing, but then the capabilities of the sensor are being underused. Considering the cost of an automotive LiDAR, approximately $1000, is too expensive to use on a motorcycle. The same end goal can be reached by simply determining the distance from the motorcycle to the vehicle behind at specific time intervals. These factors make LiDAR an illogical choice for the application. The other main light-based sensor in automotive is a Time of Flight (Tof) sensor. This sensor type, also called range finders, sends out a pulse of light and times how long it takes for the light to return to the sensor. From the time of flight measurement, the distance between the sensor to the object can be calculated. Taking these measurements on defined time intervals will provide the necessary information to calculate the velocity of the object and then acceleration. The ToF sensor will only give us the distance between the motorcycle and vehicle behind it; no excess sensor data that will need to be processed before it is ignored. The range finder Time of Flight sensor will best fit our needs at the lowest cost.