Detection and Tracking of Environmental Sensing System for Construction Machinery Autonomous Operation Application

Area of Science:

• Robotics and Automation
• Computer Vision
• Sensor Fusion

Background:

• Construction machinery operates in unstructured environments with unique sensing challenges.
• Conventional automotive sensing systems are inadequate for construction machinery's specific needs.
• Autonomous operation requires robust environmental perception for safety and efficiency.

Purpose of the Study:

• To develop and validate a LiDAR-based environmental sensing algorithm tailored for construction machinery.
• To enable real-time target detection, trajectory tracking, and prediction for dynamic objects.
• To integrate the sensing system with machine interfaces for safety functions like early warning and emergency braking.

Main Methods:

• Utilized LiDAR technology for environmental data acquisition.
• Developed algorithms for real-time target detection and dynamic object trajectory tracking.
• Implemented and tested an excavator platform for evaluating the sensing system's effectiveness.

Main Results:

• Achieved real-time detection and tracking of environmental targets.
• Demonstrated successful information exchange for safety warnings and emergency braking.
• Verified the superiority of the optimized detection model across various operating conditions and speeds.

Conclusions:

• The proposed LiDAR-based sensing algorithm effectively addresses the challenges of autonomous construction machinery operation.
• The system enhances safety through reliable environmental perception and timely decision-making support.
• Successful integration with machine interfaces enables critical safety functions, paving the way for wider adoption of autonomous construction equipment.