Navigating autonomous ground robots through un-structured and uneven terrains without relying on prior maps poses significant challenges due to irregular slopes, occlusions, and stability constraints. Path planning algorithms such as Agoraphilic* employ fixed force shaping functions to influence the robot’s movement, but these static strategies can lead to inefficient paths or trapping in cluttered environments. To address these limitations, this paper presents a novel dynamic force shaping method that adapts in real time to the robot’s local free space distribution. The proposed method utilizes a bell-shaped membership function whose form is dynamically adjusted using two parameters derived from the surrounding free space forces. When the path to the goal is unobstructed, the function sharpens to focus the navigation force toward the goal. Conversely, in constrained or cluttered areas, the function broadens to bias movement toward alternative, safer escape routes. Experiment results across diverse terrain scenarios confirm that the dynamic shaping method enhances both efficiency and robustness, enabling safer and more adaptive mapless navigation. This advancement significantly strengthens the Agoraphilic* algorithm applicability in complex real-world environments.

This paper introduces a novel path planning algorithm, Agoraphilic*, designed to address the challenges in uneven terrain environments. Unlike the traditional Agoraphilic algorithm, which is limited to navigating in 2D planes, the new algorithm extends the traditional Agoraphilic algorithm capabilities to multi-planar terrains. Building upon the basic principles of the Agoraphilic algorithm, Agoraphilic* utilizes free space searching and generates attractive forces based on available free spaces and the goal direction. While the traditional Agoraphilic algorithm estimates free spaces using distance data captured from 2D plane distance sensors, such as 2D LiDAR or ultrasonic sensors, Agoraphilic* algorithm estimates free spaces based on terrain profiles captured from 3D LiDAR or depth cameras. This adaptation equips the algorithm to navigate effectively in uneven terrain environments. By redefining the traditional Agoraphilic algorithm’s basic stages based on terrain profiles, Agoraphilic* algorithm facilitates local path planning in multi-terrain environments. The effectiveness of the proposed algorithm was validated through computer simulation.

This research presents a novel approach to the Agoraphilic local path planning algorithm by integrating semantic segmentation for enhanced terrain identification and free space navigation in complex environments. Traditional Agoraphilic algorithms, although effective in free space navigation, often struggle to accurately identify untraversable terrains such as mud or water, mistakenly considering them as navigable in ground robots. By leveraging a self-trained YOLOv8-seg network for semantic segmentation, our method identifies and labels traversable free spaces, such as roads, grass planes, footpaths and sand areas, using image input. This enhancement is applied across the core modules of the traditional Agoraphilic algorithm, resulting in a novel approach for effective navigation across challenging terrain conditions. The proposed semantic segmentation-based free space identification method is experimentally tested in real-world environments, and simulation tests validate the effectiveness of the improved Agoraphilic algorithm. This advancement represents a significant improvement in autonomous robot navigation, particularly in challenging and uneven terrains.

This paper presents a novel machine learning-based Agoraphilic (free space attraction) navigation algorithm. The proposed algorithm is capable of undertaking local path planning for mobile robots in unknown dynamic environments with a moving goal. The inability to track and reach a moving goal is one of the common weaknesses of most existing navigation algorithms operating in dynamic environments. High uncertainty involved in dynamic environments is also another major challenge. The novel machine learning-based approach helps the proposed algorithm to successfully overcome these challenges. This paper also introduces the integrated modular-based architecture for free-space attraction-based algorithms. This allows the algorithm to incorporate ten different modules with miscellaneous algorithms to perform sub-tasks such as tracking, prediction, map generation, machine learning-based free space attraction force generation and robot motion command generation. The new modular-based architecture integrates those sub-modules to create the robot's driving force. This driving force is the single attractive force to pull the robot towards the moving goal via current free space leading to future free space passages. The proposed algorithm was experimentally tested under a dynamic environment. The experiment was focused on testing the behaviour of the algorithm under the challenge of reaching a moving goal. Furthermore, the test results demonstrate that the Agoraphilic algorithm is successful in reaching a moving goal in an unknown dynamically cluttered environment. © 2022 IEEE.

This paper presents a summary of research conducted in performance improvement of Agoraphilic Navigation Algorithm under Dynamic Environment (ANADE). The ANADE is an optimistic navigation algorithm which is capable of navigating robots in static as well as in unknown dynamic environments. ANADE has been successfully extended the capacity of original Agoraphilic algorithm for static environment. However, it could identify that ANADE takes costly decisions when it is used in complex dynamic environments. The proposed algorithm in this paper has been successfully enhanced the performance of ANADE in terms of safe travel, speed variation, path length and travel time. The proposed algorithm uses a prediction methodology to estimate future growing free space passages which can be used for safe navigation of the robot. With motion prediction of moving objects, new set of future driving forces were developed. These forces has been combined with present driving force for safe and efficient navigation. Furthermore, the performances of proposed algorithm (Agoraphilic algorithm with prediction) was compared and benched- marked with ANADE (Without predication) under similar environment conditions. From the investigation results, it was observed that the proposed algorithm extends the effective decision making ability in a complex navigation environment. Moreover, the proposed algorithm navigated the robot in a shorter and quicker path with smooth speed variations.