Automated enforcement of bus-only lanes and crossings with policy and implementation insights from a Calgary case study

• Proposes cost-effective framework for bus-only lane and crossing enforcement. • Combines Raspberry Pi, GPS, and LiDAR for automated violation detection. • Uses YOLOv8 and PaddleOCR for vehicle, plate, and pedestrian recognition. • Field-tested in Calgary on 18 bus lanes and one bus-only crossing corridor. • Provides policy insights for scalable, reliable transit priority enforcement. With further attention to improving public transit service quality and reliability, transit priority measures such as bus-only lanes and bus-only crossings are becoming more common. These measures are only effective if priority can be maintained for transit, and access for regular vehicles is restricted. This can be only achieved by regulating their usage and enforcing these regulations. However, enforcement measures to prevent unauthorized vehicles from using bus-only lanes and bus-only crossings have been a manual process or employed costly systems such as bus traps in many cities. Automating this enforcement process through digital technologies can lead to substantial benefits and cost savings. Nevertheless, academic and professional literature on automated enforcement of bus-only lanes and bus-only crossings using advanced image and video processing techniques is limited. In this study, we propose a framework and guidelines for these automated enforcement systems that can be implemented by any transit agency at a low cost. We outlined our framework using three components: Hardware proof of concept, software design containing an image processing model, and user interface. We used a Raspberry Pi for hardware, PaddleOCR for license plate recognition and YOLOv8 for training our model. The operation of the system was further optimized using GPS (for bus-only lanes) and LiDAR (for bus-only crossings) sensors. To assess the applicability of our framework, we tested it on several bus-only lanes and a bus-only crossing in Calgary. The results showed a cost-effective solution while providing good performance in detecting violations and identifying unauthorized vehicles and pedestrians. The observations presented in the study can provide valuable insights to any transit agency for future implementations and policy-making process of automated bus-only lanes and bus-only crossings.