In a recent article published in the Journal of Robotics and Mechatronics, researchers explored the potential of using simple mobility algorithms to efficiently explore spaces under debris in search and rescue (SAR) operations.
They focused on developing strategies for multiple mobile robots to navigate complex and dynamic environments, ultimately aiming to locate victims trapped under collapsed buildings.
Study: Simplifying Search and Rescue Using Adaptive Algorithms. Image Credit: hxdbzxy/Shutterstock.com
Background
In the critical time of disaster like an earthquake, the initial rescue operations are essential for saving lives. Mobile robots are increasingly recognized as valuable tools in SAR efforts, providing several advantages: information gathering, victim detection, communication, and delivery of supplies.
These robots can penetrate areas under debris, providing essential details about trapped victims/individuals and assessing the structural integrity of collapsed buildings.
However, operating under debris presents significant challenges for these robots. Debris can interfere with sensors, communication, and mobility, making traditional navigation techniques like simultaneous localization and mapping (SLAM) ineffective. This highlights the need for developing robust and adaptable algorithms capable of functioning with limited information in dynamic environments.
About the Research
In this paper, the authors developed a strategy for exploring spaces under debris using multiple mobile robots. Their goal was to address the challenges posed by complex algorithms, like SLAM, by investigating the potential of simpler, primitive mobility algorithms. These simple algorithms, such as random walk (RW) and depth-first search, rely heavily on minimal environmental information, making them more adaptable to the dynamic and unpredictable conditions found under debris.
To evaluate these algorithms, the researchers designed a simulated three-dimensional (3D) environment with varying debris densities using Unity. They defined parameters such as generating obstacle area (GOA) and generated volume space (GVS) to simulate realistic debris configurations.
The debris was represented as a collection of small, randomly generated obstacles interacting with each other and the robots through physical forces like gravity and collisions. The robots, called agents, were equipped with sensors to detect obstacles and map their surroundings.
The study also evaluated the performance of individual primitive algorithms and the effectiveness of combining them using priority functions. These functions dynamically selected the most suitable algorithm based on the local environment, allowing the robots to adapt to changing conditions. Additionally, realistic simulations were conducted to test the algorithms in practical applications.
Research Findings
The outcomes showed that deterministic algorithms such as depth-first search (DFS) were effective in single-agent scenarios.
In contrast, stochastic algorithms like RW significantly outperformed them in multi-robot environments. DFS, with its systematic approach, efficiently explored spaces under debris when a single robot was involved. However, it struggled with deadlocks, where the robot became trapped and unable to proceed.
In contrast, RW's random movement pattern proved highly effective in multi-robot scenarios. The stochastic nature of RW enabled robots to explore spaces more efficiently, avoiding redundant exploration and escaping deadlocks. This efficiency was due to the robots' ability to take different paths and avoid getting stuck in the same areas.
The study also examined the effectiveness of combining DFS and RW through priority functions. A function that prioritized RW based on the local environment, called the affinity-based selection function (ASF-g), achieved the highest area coverage and exploration efficiency in multi-robot scenarios.
The hybrid approach demonstrated enhances the exploration process's efficiency and effectiveness. Additionally, the simulations showed that the developed methodology, which includes generating diverse debris scenarios, offers valuable insights into the practical performance of these algorithms.
Applications
This research highlights significant advancements in improving the effectiveness of SAR operations. By employing simple, adaptable algorithms such as RWs, rescue teams can deploy multiple robots to efficiently navigate complex and dynamic environments under debris in the disaster site.
This approach can speed up the identification of the location and condition of trapped victims, thereby improving their chances of survival.
Efficient exploration of disaster sites allows for optimal resource allocation, enabling rescue teams to concentrate efforts on areas where victims are most likely to be found. Utilizing robots to explore dangerous areas can reduce risks to human personnel, ensuring their safety.
Additionally, robots can facilitate two-way communication between victims and rescue teams, providing essential information and comfort. They can also deliver vital supplies, including food, water, and medical aid, further increasing the likelihood of survival for those trapped.
Conclusion
The paper demonstrated that primitive mobility algorithms and novel strategies improved exploration in SAR scenarios. It showed that stochastic algorithms were effective in multi-robot settings, especially in dynamic and unpredictable environments.
The findings suggested that combining simple algorithms with adaptive priority functions could significantly enhance the efficiency of robot exploration in complex disaster situations.
Future work should focus on improving collaboration among multiple robots. This included developing communication strategies to share information about unexplored areas and coordinate efforts, potentially increasing efficiency.
The authors also recommended integrating robots with various capabilities, like obstacle removal, to further enhance SAR operations.
Journal Reference
Disclaimer: The views expressed here are those of the author expressed in their private capacity and do not necessarily represent the views of AZoM.com Limited T/A AZoNetwork the owner and operator of this website. This disclaimer forms part of the Terms and conditions of use of this website.