Human-Swarm Interaction and Collaboration

A. Introduction

The number of mobile robots used in the field has significantly increased during the last few decades. Their widespread use gives the obvious benefits of lower costs, the protection of individuals from harm, or the opening up of whole new uses that were previously impractical. New jobs and missions, including search, exploration, rescue, surveillance, pursuit, and infrastructure deployment, are made possible, particularly by the integration of extremely large teams of robots into comprehensive systems. Applications span a wide range of fields, from low-cost warehouse security to search and rescue to extra-terrestrial exploration. The trend toward using large teams of mobile robots is expected to be further accelerated by new advances in commodity electronics that act as inexpensive substitutes for otherwise expensive sensor or motion capabilities. However, this presents a barrier for managing such systems, particularly for human operators. A number of operators, frequently using remote control, manages the majority of big robotic systems in use today. Such a method is not workable for larger systems with many robots and inexpensive technology. While tools like mappers, path planners, and monitoring and detecting systems are already essential for extremely powerful systems, autonomy is anticipated to become even more crucial for robotic systems with a very high number of robots, or so-called swarms. However, as autonomy is used and relied upon more, it presents a new challenge for human operators, particularly when distributed algorithms with complicated dynamics are employed. In other words, it's still a challenge for human controllers to manage robot swarms of hundreds or more. Swarms are more challenging to regulate due to their high number of robots as well as the fact that their desired functioning rely on the interaction of individual robots developing a useful attribute. One of the main characteristics of the majority of swarms is this phenomena of non-apparent emergent activity. The function of today's robotic systems is typically determined by the intricate interactions of a single, strong robot with its operator and environment. However, in most swarms, function is determined by how a large number of robots interact with one another. The emergent behaviour should, in theory, be resilient to environmental changes, robot malfunctions, and other unforeseen events and not depend on each individual robot operating as intended. The problem is that the majority of the time, the individual robot behaviour does not match a desired as depicted in figure 1.

Figure 1.

The consideration of human factors that influence human performance throughout a mission becomes imperative when human participation is included. In light of this, developing an interaction paradigm that considers human factors enhances both the execution of the mission as a whole and of individuals. A realisation of such an interaction scheme exists in flexible autonomy systems in which task distribution and interface customisation can be contingent on the state of the mission including the human. Although systems with flexible autonomy have been shown to be superior to rigid systems with fixed autonomy, many aspects of flexible autonomy are still poorly understood. In the article we begin in section A with the introduction, The literature review is discussed in section B, In section C we are discussed about different Swarm Cluster Model. Human and Swarm interaction types is discussed in section D. Cognitive Complexity of Human-Robot System is discussed in section E followed by next research and summary in section F.

B. Literature Review

Bayındır (2016) worked on distributed algorithms are shown to bring cooperation between agents, obtained in various forms and often without explicitly programming a cooperative behaviour in the single robot controllers. Offline and online learning approaches are described, and some examples of past works utilizing these approaches are reviewed.

Brown et al., (2014) uses a swarm model with two attractors, we demonstrate this concept by showing how limited human influence can cause the swarm to switch between attractors. We further claim that using quorum sensing allows a human to manage trade-offs between the scalability of interactions and mitigating the vulnerability of the swarm to agent failures.