Swarm Intelligence Workshop on Distributed Robotics: From Groups to Swarms
Abstract
There are many challenges in current Swarm Intelligence (SI) research [1], from biological inspiration to algorithm development andreal-world implementations. Multi-robot systems in general, from groups to swarms [2,3], will be the main topic of the workshop from technological and scientific perspectives. Challenges include improving our understanding of swarm concepts such as coordination and decision-making in swarms, resolving uncertainty in swarms through collective and multisensory perception, and different approaches to information sharing and processing in swarms [4]. The ability to deploy robot swarms is severely constrained by the currently limited capabilities of autonomous robots. In order for a collection of autonomous robots to form a swarm, the individual robots must be able to interact and communicate with each other, as well as recognize their peers and the work they bring to the collective. This requires tailored hardware designs, specific sensing and processing, anda variety of interaction capabilities. Swarm robotics applications often extract engineering principles from the study of natural systems to provide multi-robot systems with similar characteristics and capabilities. The goal is to build systems that are more robust, fault-tolerant, and flexible than individual robots, with the ability to better adapt their behavior to changes in the environment. Emerging issues include how to move from a microscopic level, i.e., individual robot behavior, to the desired macroscopic swarm behavior, and vice versa; how to bridge the gap between simulation and real-world experimentation, and between laboratory experiments and real-world domains. Apart from robotics, the exploration of SI applications as used in data science and data mining or operations research [5, 6] is of interest. This approach is motivated by recent algorithmic developments that show how SI has been successfully used to solve complex optimization problems by simulating self-organizing mechanisms inherent in biological swarms. In data science applications [7], the goal is to develop algorithms for automatically extracting knowledge from immense amounts of data. Another study [8] proposes to use SI to drive data mining parameter tuning/optimization criteria in machine learning and statistics. Moreover, in next generation wireless communication networks, a very large number of devices and applications are emerging in the context of SI, along with the heterogeneity of technologies, architectures, mobile data, and the need for novel methods for network optimization, including spectrum management and resource allocation, wireless caching, and edge computing, among others [9]. The workshop will address new and emerging concepts in SI in a multidisciplinary way, with this edition focusing on distributed robotics aspects, as it relates to how biological entities in nature self-organize and exhibit collective behavior to cooperatively solve problems or perform tasks in real-world challenges, e.g., wide area search in unstructured complex environments.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Oct 13, 2023
- Source ID
- N629092312092
Entities
People
- Alcherio Martinoli
Organizations
- Office of Naval Research
- Swiss Federal Institute of Technology in Lausanne
- United States Navy