BikeSight is an integrated system aiming to prevent and respond to dooring biking accidents. It was designed and prototyped by students of the Master in Design Engineering program at Harvard University. The solution leverages machine vision, edge computing, and IoT technology to prevent common dooring accidents in urban and suburban settings. It is a modular product attached to a car's rearview mirror and connecting to its infotainment and locking systems. 

As a car is parked, BikeSight prevents dooring accidents by identifying incoming cyclists and automatically locking the car's doors. It uses Google's TensorflowLite deep learning technology to identify a cyclist's speed and position. Trained models leverage OpenMV microcontrollers and use edge computing technology to detect accidents without the need for an internet connection. In the event of an accident, BikeSight can automatically call emergency services and provide operational guidance to users through Apple CarPlay or Android Auto.

Human-centred design was core to the development of BikeSight. During ideation and prototyping, user interviews with drivers and cyclists allowed the identification and prioritization of opportunities. In parallel, the design team also researched the legal framework relevant to bicycle accidents. The  BikeSight team collaborated with a personal injury lawyer to ensure that the system provides adequate support to users who might not know what steps to take in the event of a crash.