AI-Assisted Self-Powered Vehicle-Road Integrated Electronics for Intelligent Transportation Collaborative Perception

Area of Science:

• * Intelligent Transportation Systems
• * Autonomous Driving Technologies
• * Embedded Systems

Background:

• * Limited perception capabilities of current autonomous driving systems necessitate enhanced sensing solutions.
• * Vehicle-road interaction is crucial for improving safety and efficiency in intelligent transportation.
• * Existing data analysis methods for vehicle-road data often lack sufficient accuracy.

Purpose of the Study:

• * To design and validate self-powered vehicle-road integrated electronics (SVRIE) for collaborative perception.
• * To investigate the dual role of SVRIE as both a tire-integrated device and an embedded road surface array.
• * To assess the performance of SVRIE in monitoring vehicle motion, road conditions, and tire health.

Main Methods:

• * Development of SVRIE with a multilevel fractal structure for pressure sensing and anti-crosstalk.
• * Characterization of SVRIE array performance for cooperative vehicle-infrastructure systems.
• * Application of convolutional neural networks (CNNs) and transfer learning for data analysis and classification.

Main Results:

• * SVRIE demonstrates high performance in collaborative sensing of vehicle motion, road surface conditions, and tire health.
• * CNN model achieved at least 88.3% accuracy in classifying vehicle and road conditions.
• * Transfer learning model reached 100% accuracy for road surface identification, 97% for tire motion recognition, and 99% for tire health monitoring.

Conclusions:

• * SVRIE technology offers a novel approach to enhance collaborative perception between vehicles and road infrastructure.
• * The proposed system significantly improves the accuracy of monitoring critical driving parameters.
• * This research provides a foundation for future advancements in intelligent transportation and autonomous driving safety.