Improving Pedestrian Safety Using Ultra-Wideband Sensors: A Study of Time-to-Collision Estimation

Area of Science:

• Traffic Safety Engineering
• Sensor Technology
• Human-Computer Interaction

Background:

• Pedestrian safety is often assessed by collision data, but traffic conflicts offer more frequent insights.
• Current video-based conflict detection is limited by weather and lighting.
• Wireless sensors offer a robust alternative for traffic conflict monitoring.

Purpose of the Study:

• To develop and evaluate a prototype safety assessment system using ultra-wideband (UWB) wireless sensors.
• To detect traffic conflicts and provide real-time collision avoidance alerts.
• To assess the system's performance in adverse conditions.

Main Methods:

• Utilized ultra-wideband wireless sensors for traffic conflict detection.
• Implemented a customized time-to-collision algorithm for conflict severity assessment.
• Conducted field trials simulating vehicle and pedestrian sensors, calculating proximity in real-time.

Main Results:

• Demonstrated the feasibility of using UWB sensors for real-time traffic conflict detection.
• Validated the accuracy of time-to-collision measurements.
• Showcased the system's ability to alert smartphones and prevent collisions in various conditions.

Conclusions:

• UWB wireless sensors provide a robust solution for traffic conflict detection, augmenting existing methods.
• The developed system enhances pedestrian safety by offering real-time alerts, irrespective of weather.
• Further research is recommended to address identified limitations and improve system capabilities.