Edge Detection in Static and Dynamic Environments using Robot Swarms

Abstract

Abstract-Robotic systems offer an attractive solution for a large spectrum of real-world applications that are hosted in dangerous or inaccessible areas for humans. In such applications, one of the fundamental tasks is to detect and mark particular features (e.g., pollution areas) in order to develop a proper response. In this paper, we focus on the specific problem of environmental edge detection using a swarm of homogeneous robots that can sense and act distributively using a large number of individuals. In our study, we consider both static and dynamic 2D environments, in which the spatial distribution of the feature(s) may change over time. We verify our results using a set of physics-based simulations.

Construction Task Allocation Through the Collective Perception of a Dynamic Environment

“…In this study, we tackle a specific process of collective decision-making that is referred to as collective perception. Collective perception is a process, in which the robots are engaged in perceiving an environmental stimulus collaboratively [30,36,14,31]. In general, robots need to perceive particular signals in their environments to make decisions based on the perceived values.…”

Multi-Functional Sensing for Swarm Robots Using Time Sequence Classification: HoverBot, an Example

“…The work presented here has the potential to enhance a robot's sensing capabilities without modifying its hardware, hence, could add to the list of observable features for collective perception. Notable literature on collective perception includes Khaluf 's work on detecting and marking features e.g., of pollution areas (Khaluf, 2017)…”

“…The work presented here has the potential to enhance a robot’s sensing capabilities without modifying its hardware, hence, could add to the list of observable features for collective perception. Notable literature on collective perception includes Khaluf’s work on detecting and marking features e.g., of pollution areas ( Khaluf, 2017 ), Kornienko et al’s work on investigating which sensing and processing steps should be done individually or collectively for collective perception with robot swarms ( Kornienko et al, 2005a ), Schmickl et al’s work on hop-count and Trophallaxis-inspired strategies to collectively perceive targets ( Schmickl et al, 2007 ), Mermoud et al’s work on aggregation-based strategies to collectively perceive and destroy specific targets ( Mermoud et al, 2010 ), and Tarapore et al’s work on collective perception strategies inspired by the adaptive immune response to discriminate between dangerous and friendly cells ( Tarapore et al, 2013 ).…”

Collective Lévy Walk for Efficient Exploration in Unknown Environments

“…Exploring unknown environments to spot targets is one of the most fundamental problems in the context of mobile robots used for search and rescue, environment mapping or agricultural applications [3]. An efficient exploring strategy that provides a maximized area coverage in a minimized time interval is the main design goal.…”