University of Michigan

Dongyao Chen

Kang Geun Shin

Kyong-Tak Cho

Left turns are known to be one of the most dangerous driving maneuvers. An effective way to mitigate this safety risk is to install a left-turn enforcement—for example, a protected left-turn signal or all-way stop signs—at every turn that preserves a traffic phase exclusively for left turns. Although this protection scheme can significantly increase the driving safety, information on whether or not a road segment (e.g., intersection) has such a setting is not yet available to the public and navigation systems. This disclosure presents a system that exploits mobile crowdsensing and deep learning to classify the protection settings of left turns.

Inferring left-turn information from mobile crowdsensing

Driver fingerprinting using sensor data was known to be feasible only with access to in-car data. This disclosure presents a novel technique for identifying a vehicle driver from only one vehicle turn and using zero-permission sensors residing in the mobile device. Through extensive evaluations, extracted features are shown to reflect only the drivers unique turning style and thus functions as the core of driver fingerprinting.

Fingerprinting drivers based on vehicle turns

A vehicle maneuver detection application is proposed for driving assistant systems. The application can accurately and inexpensively detect and differentiate vehicle steering maneuvers by utilizing built-in sensors on smartphones or other portable computing device residing in a vehicle. By leveraging an effective bump detection algorithm and studying the nature of steering, the application is capable of differentiating various steering patterns, such as lane change, turn, and driving on curvy roads. Practicality of the application is demonstrates by two use cases: careless steering detection and fine-grained lane guidance. Thus, the application provides new functionalities without relying on cameras to provide a broader range of driving assistance.

Growing community of inventors

Ann Arbor, MI, United States of America

Dongyao Chen