Proposed System: Design Pillars • Develop an intelligent Intersection Safety System (ISS) architecture to increase safety of all road users: Improving awareness of road users via visuo-audio cues and controlling intersection controls, e.g. turning all traffic-lights to red, can prevent accidents from happening. • Employ data fusion with existing and emerging sensors: Using low-cost sensors and utilizing already existing devices to get road user data and create redundancy for safety. • Augment data processing with AI/ML: Using mature deep learning based object detection algorithms and pattern recognition pipelines can improve awareness and anticipate potential conflicts. • Design a low-cost, high-impact deployment strategy: Integration with already existing intersection infrastructure to increase safety and enable rapid widespread deployment. • Use ISS as a data collection platform: Data recording and post processing for continuous learning, identifying high-risk patterns, and crash reporting. Data analytics obtained through the proposed ISS can support better intersection geometry design for improved safety.
122
Made with FlippingBook Annual report maker