Range-Finding Rear-Collision Accident Prediction and Warning For Motorcycles with Integrated Signal Vest
The project goal is proof of concept, that a cheap sensor can be developed and applied to motorcycles. Despite the fact that the project was not complete in the sense of having a working accident prediction system, it still proves that a sensor can be made for cheap. The total cost of all the parts used in the project amounts to $84. Compared to the $1k to $3k that a LIDAR system costs, this is astronomically cheaper. The parts were chosen based on their price and effectiveness. There are parts out there that met the parameters better than the ones used, but would drastically increase the price of the system. An example of this is the TIA op amp used. The op amp that fit the job the best is a $300 high-speed LIDAR geared TIA op amp. This would have made a big difference in the system, but costs too much. This is how cost affected the development of this project.
With the received signal, only a few more months of engineering is needed to finish the sensor portion. Regardless, the fact that there was a detectable time difference between the two distances measured proves that the sensor can work. In this aspect, the project is a success. It has been proven that a cost effective sensor is designable. The LED display also returned positive results. The model shows that the grid will be bright enough to be used at all times and is able to depict the different signals. For these reasons, the project is a success.
As stated before, there is still work to be done. The sensor system needs to be completed. The noise from the MOSFET driver needs to be eliminated, and the resulting signal needs to be cleaned up. Cleaning up the signal will require the use of lenses and filters. The optical focusing lenses will also help the system achieve the full range required. The filters will help eliminate the external light noise. Though the affect of external light was not determined to be a problem in the lab, there is still a large possibility for light noise to distort the signal. The tests conducted were performed in a "dark" (no lights turned on) lab. The system has not been tested outside in the sunlight or under road lights. The accident prediction algorithms also need to be implemented to the system. The algorithms will need to be application specific due to the timing of the sensor. The last step for the project is to make the sensor on a PCB. This will solve several problems experienced during design. The voltage level issues between the MCU and the TDC7200 will be fixed since the MCU can be upgraded to a surface mount MCU with a higher clock rate and 3.3V power supply. The thermal overheat potential for the power conversion can be mitigated with the use of the thermal pad.
The project as a whole is a success considering its ability to complete the proof-of-concept goal. The next steps for the project have been clearly laid out, and once completed, the motorcycle community will be a little bit safer.