On-the-Go Path Planning and Repair in Static and Dynamic Scenarios

TL;DR

The effectiveness of the proposed dynamic, real-time path planning approach is demonstrated, demonstrating its suitability for robotic path planning in both known and unknown environments, including those involving mobile objects, agents, or potential threats.

Abstract

Autonomous systems, including robots and drones, face significant challenges when navigating through dynamic environments, particularly within urban settings where obstacles, fluctuating traffic, and pedestrian activity are constantly shifting. Although, traditional motion planning algorithms like the wavefront planner and gradient descent planner, which use potential functions, work well in static environments, they fall short in situations where the environment is continuously changing. This work proposes a dynamic, real-time path planning approach specifically designed for autonomous systems, allowing them to effectively avoid static and dynamic obstacles, thereby enhancing their overall adaptability. The approach integrates the efficiency of conventional planners with the ability to make rapid adjustments in response to moving obstacles and environmental changes. The simulation results discussed in this article demonstrate the effectiveness of the proposed method, demonstrating its suitability for robotic path planning in both known and unknown environments, including those involving mobile objects, agents, or potential threats.

Figures (12)

• Figure 1: The static obstacle on the right necessitates the planner to circumvent with a longer path. In contrast, the left randomly appearing obstacles within a compact space, supposedly occupied by a larger obstacle, allow the planner to provide a more optimal path.
• Figure 2: The top-left section illustrates the 2D environment, while the top-right and bottom sections represent the two 3D environments. In these diagrams, the purple points mark the agent's starting position, and the green points denote the target location. The red bars represent the obstacles present in each environment.
• Figure 3: An illustration of the Planner on the static 2D environment as the agent progresses through successive steps until reaching $x_{goal}$.
• Figure 4: An illustration of the Planner on the dynamic 2D environment evolving according to Case $1$ as the agent advances through each step until reaching $x_{goal}$.
• Figure 5: An illustration of the Planner on the dynamic 2D environment evolving according to the first scenario in Case $2$ as the agent progresses step by step until reaching $x_{goal}$.