DSSE: a drone swarm search environment

TL;DR

The paper presents DSSE, an open-source, PettingZoo-based environment for simulating drone swarm search tasks in which a shipwrecked person follows a probabilistic, drift-driven movement on a grid. It introduces a probability matrix that evolves with ocean current and uncertainty, and a reward scheme that guides agents to explore high-probability areas efficiently. The work emphasizes reproducibility and public accessibility, providing documentation and a Python package to support RL research in dynamic probability inputs. Real-world scaling considerations tie the grid to practical sensor specifications, while acknowledging current simplifications and outlining avenues for future improvements.

Abstract

The Drone Swarm Search project is an environment, based on PettingZoo, that is to be used in conjunction with multi-agent (or single-agent) reinforcement learning algorithms. It is an environment in which the agents (drones), have to find the targets (shipwrecked people). The agents do not know the position of the target and do not receive rewards related to their own distance to the target(s). However, the agents receive the probabilities of the target(s) being in a certain cell of the map. The aim of this project is to aid in the study of reinforcement learning algorithms that require dynamic probabilities as inputs.

Figures (3)

• Figure 1: Map representation
• Figure 2: Initial state of a probability matrix (color represents probability)
• Figure 3: Probability matrix after the time (color intensity represents higher probabilities)