OA-Bug: An Olfactory-Auditory Augmented Bug Algorithm for Swarm Robots in a Denied Environment

Siqi Tan, Xiaoya Zhang, Jingyao Li, Ruitao Jing, Mufan Zhao, Yang Liu, Quan Quan

TL;DR

This paper introduces OA-Bug, a novel algorithm that equips swarm robots with olfactory and auditory sensors to navigate and explore "denied environments" (where GPS, mapping, and communication are unavailable). By mimicking animal-like cooperation, OA-Bug significantly improves search coverage compared to existing methods, demonstrating its effectiveness in both simulation and real-world experiments.

What Problem Does This Paper Address?

Imagine a disaster scenario where communication is down, GPS signals are lost, and robots can't rely on pre-existing maps. This is a "denied environment," and it poses a significant challenge for robotic search and rescue (RSAR) operations. Traditional robotic navigation often relies on these external aids. While some algorithms exist for multi-robot systems (MRS) in such conditions, they often require internal memory, can be computationally expensive, or depend on centralized systems that might fail. The core problem is enabling a swarm of robots to efficiently and cooperatively explore an unknown, unstructured environment without any external positioning or communication support, maximizing their chances of finding victims or covering the area.

Methods

The researchers propose the Olfactory-Auditory Augmented Bug Algorithm (OA-Bug). This algorithm builds upon the traditional "Bug Algorithm," which uses local sensing and a general direction to navigate. OA-Bug enhances this by integrating two key sensory modalities:

  • Olfactory Sensing (Ethanol Scent): Each robot leaves a trail of ethanol as it moves. This scent acts as a marker, allowing other robots to detect if an area has already been visited. This helps reduce revisits and improves exploration efficiency.
  • Auditory Sensing (Bluetooth AoA): Robots can emit signals that other robots can detect using Angle of Arrival (AoA) technology. This allows robots to determine the relative positions of their swarm mates, aiding in coordination and helping them choose better directions when encountering obstacles or deciding on new exploration paths.

The algorithm operates using a finite state machine with states like "Just Started" and "Following Wall." Key triggers include hitting walls or other robots, and detecting visited areas via olfactory cues. The auditory information is crucial for updating a robot's "preferred direction," which is not tied to a specific target but rather a general direction for exploration, allowing for more dynamic and adaptive behavior.

Datasets

Not discussed in the paper. The authors developed their own simulation environment and conducted experiments on real robots in a controlled setting.

Key Insights and Ideas

  • Nature-Inspired Cooperation is Key: The paper strongly suggests that mimicking how animals cooperate using simple sensory cues (like scent trails or calls) is a powerful strategy for robots in denied environments.
  • Olfactory + Auditory Synergy: Combining scent marking (for visited areas) with directional auditory cues (for relative positioning) creates a robust system that significantly outperforms algorithms relying on only one of these or neither.
  • Adaptive Exploration: The "preferred direction" concept, which can be updated by auditory cues, allows robots to dynamically adjust their exploration strategy, preventing them from getting stuck in local optima or repeatedly visiting the same areas.
  • Reducing Revisits is Crucial for Coverage: The olfactory sensor's ability to detect previously visited areas is a primary driver of the algorithm's high coverage rates.
  • Departure Point Matters: Robots departing from the center of a site tend to spread more evenly and achieve higher coverage with fewer actions compared to departing from the boundary, though the algorithm is designed to perform well in both scenarios.

Key Results

  • Simulation: OA-Bug achieved a search coverage of 96.93% in simulations, significantly outperforming a similar algorithm (SGBA). Algorithms relying solely on olfactory or auditory cues performed much worse, highlighting the importance of their combination.
  • Real-World Experiments: In real-world tests with four robots, OA-Bug covered an average of 15.167 out of 18 rooms, resulting in 84.26% coverage. This was achieved despite the challenges of real-world environments (e.g., non-flat walls) and occasional hardware failures.
  • Efficiency: The algorithm demonstrated efficient exploration, often requiring fewer actions to achieve high coverage compared to other tested methods.

Limitations and Future Work

  • Limitations:
    • The current implementation assumes idealistic sensors without uncertainties.
  • Real-world environments present challenges like non-flat walls and potential hardware failures, which can impact performance.
  • The current focus is solely on searching; victim detection is reserved for future work.
  • Future Work:
    • Applying OA-Bug to more complex tasks beyond searching, such as returning to a departure point, handling low power, or responding to summons.
  • Analyzing and optimizing the algorithm by considering macro-level behavior rather than just micro-level details.
  • Developing more precise sensor models.
  • Porting the system to Unmanned Aerial Vehicles (UAVs).

Summary updated

Abstract

Searching in a denied environment is challenging for swarm robots as no assistance from GNSS, mapping, data sharing, and central processing is allowed. However, using olfactory and auditory signals to cooperate like animals could be an important way to improve the collaboration of swarm robots. In this paper, an Olfactory-Auditory augmented Bug algorithm (OA-Bug) is proposed for a swarm of autonomous robots to explore a denied environment. A simulation environment is built to measure the performance of OA-Bug. The coverage of the search task can reach 96.93% using OA-Bug, which is significantly improved compared with a similar algorithm, SGBA. Furthermore, experiments are conducted on real swarm robots to prove the validity of OA-Bug. Results show that OA-Bug can improve the performance of swarm robots in a denied environment. Video: https://youtu.be/vj9cRiSm9eM.
BibTeX
@misc{tan2022oa,
  author = {Siqi Tan and Xiaoya Zhang and Jingyao Li and Ruitao Jing and Mufan Zhao and Yang Liu and Quan Quan},
  title = {{OA-Bug: An Olfactory-Auditory Augmented Bug Algorithm for Swarm Robots in a Denied Environment}},
  year = {2022},
  eprint = {2209.14007},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2209.14007},
  howpublished = {arXiv},
}
Figure