ECE.14 – Vehicle Blind Spot Detection System
Team Members Heading link
- Raul Costilla
- Edward Machado
- Rashell Uruchima
Project Description Heading link
Each year thousands of people are killed or seriously injured as a result of a vehicular accident. Research shows that human factors account for 80 percent of these accidents. Injuries resulting in an accident are more common in some types of vehicles than others. For example, motorcyclists are 29 times more likely to die in a crash than passengers in any other type of vehicle. Speed and failure to yield are primary causes of accidents. Lane changes are also a key contributor to accidents. In spite of this, some vehicles lack basic safety features that address these issues. For example, while many high priced vehicles may have blind spot detection systems, many more modestly priced vehicles lack sensors that can identify such hazards. We have developed a blind spot detection system for such vehicles. The blind spot detection system can be retrofitted to any vehicle e.g. automobile, motorcycle, bicycle, etc. This solution uses recently deployed small form factor mmWave technology which has a proven performance track record of detecting potential hazards in a variety of driving environments such as rain and smog. The current implementation utilizes an antenna-on-chip or AOP phased array with mmWave based radar that performs object detection behind the rider, with adjustable sensitivity and a programmable field of view. In our implementation the mmWave AOP communicates information via USB to a Raspberry Pi micro computing system. A GATT server that we implemented on the Raspberry Pi communicates this information via Bluetooth to a custom iOS application we wrote that runs on a smartphone. The application provides an indicator to the driver alerting them of the presence of objects behind the vehicle to the left and right side of the driver. Our project will indicate if the potential hazard is on the back left or the back right side of the driver.