Guidance Point Generation-Based Cooperative UGV Teleoperation in Unstructured Environment

Area of Science:

• Robotics
• Artificial Intelligence
• Autonomous Systems

Background:

• Teleoperation of unmanned ground vehicles (UGVs) is crucial in military and civilian applications.
• Current teleoperation methods are limited by low video resolution, restricted field of view, and control loop time delays, necessitating reduced vehicle speeds for stability.
• Existing systems often require predefined goal positions, limiting adaptability in unstructured environments.

Purpose of the Study:

• To propose a novel guidance point generation method for human-machine cooperative UGV teleoperation.
• To enable UGV navigation in unstructured environments without relying on predefined goal positions.
• To enhance teleoperation speed and stability by utilizing local perception information.

Main Methods:

• Generation of a locally occupied grid map (OGM) using probabilistic grid state description.
• Conversion of OGM to a binary image to construct obstacle area convex hulls.
• Application of an improved thinning algorithm to extract navigable region skeletons.
• Utilizing k-nearest neighbor (kNN) to identify the target skeleton relative to UGV position.
• Reconstruction of the target skeleton using a decreasing gradient algorithm for optimal guidance points.
• Transformation of guidance point selection to trajectory selection based on UGV's differential equation of motion.

Main Results:

• The proposed method successfully generates guidance points using only local perception data.
• The technique effectively identifies navigable paths and reconstructs skeletons for guidance.
• Experimental results demonstrate a significant increase in the speed of teleoperated UGVs.

Conclusions:

• The novel guidance point generation method enhances UGV teleoperation in complex environments.
• The approach improves human-machine cooperation by providing intuitive trajectory selection.
• This method offers a viable solution for increasing the operational speed and efficiency of teleoperated UGVs.