Approach and Fork Insertion to Target Pallet Based on Image Measurement Method.

Area of Science:

• Robotics
• Computer Vision
• Automation Engineering

Background:

• Previous research proposed image processing for pallet detection and inclination angle estimation.
• Existing methods required improvement for environmental robustness in automated logistics.

Purpose of the Study:

• To enhance the robustness of image-based navigation for autonomous forklifts.
• To achieve stable, environment-independent pallet approach and fork insertion operations.
• To demonstrate the integrated system's functionality in a simulated warehouse environment.

Main Methods:

• Improved image processing algorithms for pallet detection and pose estimation.
• Developed a robust control strategy for the approach and fork insertion phases.
• Integrated algorithms onto an Autonomous Guided Forklift (AGF) platform.

Main Results:

• The enhanced method demonstrated stable performance in various environmental conditions.
• Successful and consistent execution of the complete pallet handling sequence (approach to fork insertion) was achieved.
• The system proved effective in a simulated warehouse setting.

Conclusions:

• The improved image-based navigation system significantly enhances the robustness and stability of autonomous forklift operations.
• This advancement enables reliable automated pallet handling, crucial for warehouse efficiency and safety.
• The successful demonstration paves the way for practical deployment of AGFs in complex logistics environments.