A Smart Bicycle That Protects Itself

Abstract

This article presents a smart bicycle with inexpensive sensors, active sensing control systems, and estimation algorithms for the prevention of car-bicycle collisions. The smart bicycle tracks nearby cars, predicts their trajectories, and (in the case of the danger of a possible car-bicycle collision) sounds an audio alarm to alert the motorist to the presence of the bicycle. The proposed technology works on today's roads and does not depend on any additional technology being available on cars. The challenges in developing the proposed system stem from the inability to utilize the expensive radar and high-density lidar sensors that are typically used on prototype autonomous cars. The system is also required to address complex traffic scenarios involving rear-approaching, left-turning, and right-turning vehicles. To enable the bicycle to reliably track nearby vehicles using novel inexpensive sensors, active sensing control and estimation algorithms for vehicle tracking are developed. Human-factors research is used to create an appropriate auditory alert that attempts to convey both a sense of emergency and the resemblance to a bicycle bell. This article describes the developed technology and its implementation on prototype e-bikes. A three-week field operational test was then conducted using the prototype smart bicycles to evaluate the effectiveness of the developed technology on today's roads. The field operational test involved three instrumented bicycles, bicyclist volunteers with significant daily urban commutes, and an analysis of bicycle data recorded during their commutes in real-world traffic conditions. sensor system first tracks a vehicle that has a larger entropy of the position uncertainty. Otherwise, the laser sensor system conducts the search task based on search area uncertainties (the laser sensor first searches an area with larger uncertainty) using (S5). If two vehicles are driving in an inside adjacent lane, then only the first vehicle will be tracked by the estimation system. It is possible that there are some unique circumstances (for example, where the car behind a school bus makes a sudden lane change to directly behind the bicycle and poses a danger to the bicycle) that cannot be reliably tracked. In such a case, it is possible that the sensing system might be late in detecting the danger. However, such a danger would exist even if the bicycle rider was directly watching and not using a sensor-based automatic tracking system. The automatic sensing system would still assist the bicycle rider, who would otherwise be personally watching for all vehicles.