An Inductive Loop Detector for Personal Transporters in Urban Cycle Lanes

A sensor coil is developed for detecting personal transporters based on unique vehicle signatures. Mixed use of a wide range of personal transportation vehicles like kick scooters, skateboards and hover boards on the urban cycle lanes have led to an increased traffic on these lanes. For the planning and optimal use of such lanes, accurate information regarding the traffic on such lanes, i.e. the number and types of vehicles are needed. Most of the conventional vehicle detection schemes fail to sense these vehicles accurately due to their small size. As the vehicle size is less compared to the rider, even the computer vision based techniques confuse them as pedestrians. Inductive loop sensors have several advantages and are widely used for traffic detection of large vehicles. They provide a signature for each type of vehicle, due to the change in inductance of the sensor coil owing to the metallic bottom structure of the vehicle. The present work proposes four sensor coil designs for cycle lanes, and evaluates their performance in sensing and classifying the personal transporters. A simple coupled resonating circuit that gives high sensitivity and noise rejection capability is used to measure from the sensor coil. The sensor designs are tested with three types of vehicles, typically found on cycle lanes -electric kick scooter, bicycle and a skateboard. The vehicles are driven through various possible paths along the sensor coil and the corresponding signatures are recorded. From the observed signatures for each vehicle type, a set of attributes are identified, which could be used as inputs for any classification algorithm. The proposed designs are also evaluated based on a set of criteria, from which, one of the design is selected as the best. This study shows that, if suitably designed, inductive loop sensors can give unique signatures for personal transportation vehicles, enabling reliable detection and classification.