Distributed odor source localization

Abstract

This paper presents an investigation of odor localization by groups of autonomous mobile robots. First, we describe a distributed algorithm by which groups of agents can solve the full odor localization task. Next, we establish that conducting polymerbased odor sensors possess the combination of speed and sensitivity necessary to enable real world odor plume tracing and we demonstrate that simple local position, odor, and flow information, tightly coupled with robot behavior, is sufficient to allow a robot to localize the source of an odor plume. Finally, we show that elementary communication among a group of agents can increase the efficiency of the odor localization system performance.

Multi-robot olfactory search in structured environments

“…Most of the related works concerning olfactory search have focused on either single robot experiments [1,2,3] or multiple robots operating in open areas free of obstacles [4,5,6] with a background fluid flow. The odor is carried downwind originating from the source forming a plume.…”

“…The iRobot 4 Roomba was used in the experimental tests. This is an attractive platform because it is inexpensive, readily available and can be fully monitored and commanded through a serial port interface.…”

“…For example, the value of k (common subgraph size) in conditions (4) and (5) is a critical issue. Based on simulation experiments, it was figured out that it is good enough to set k to eight.…”

A Graph-Based Formation Algorithm for Odor Plume Tracing

“…Many algorithms take inspiration from living creatures, such as bacteria [4] or silkworm moths [20], and generally operate by switching a single robot among a set of simple behaviors. Experiments have also been conducted with multiple robots [10,17,13], acting both independently and cooperatively. Other approaches include Braitenberg-type control [16], probabilistic inference [28,17,15] and meta-heuristic optimization methods [1,2,3,12,21].…”

“…Our work is strongly inspired by the crosswind formation work in [18], as well as other formation-and swarm-based approaches [8,19,10]. This decision is driven by the choice of success metric: while the energy budget goes up with the number of robots, these techniques allow us to minimize the distance overhead and time to the source, a requirement in scenarios such as the aforementioned search and rescue.…”