Novel Swarm Behaviors Through Physical Inter-Robot Connections
Abstract
In nature, swarms have impressive behaviors to capture objects or build structures, such as towersand bridges. These behaviors relyon the physical interaction between individual agents. In thisresearch proposal, we develop novel swarm behaviors that allow aerialrobots to operate in complexenvironments relying on physical inter-robot interaction. The physical interconnection betweenaerial robots brings new theoretical challenges to the design and control of aerial vehicles. Ourobjective is to co-design robot hardware andcontrol algorithms that allow small agile robots tocooperate and generate emerging behaviors. These swarm behaviors allow navigation, conglomeration,and object transportation in turbulent environments. The complexity of the connectionsincreases along the project,starting with articulated mechanisms that considerably reduce thedimensionality of the configuration of space. Then, we explore bendable materials that involve reinforcementlearning techniques to achieve thedesired behavior. Most behaviors involve performingconglomeration or rendezvous of a subset of robots in the team. This approach reduces the complexityof the system and the required equipment for operation. To make the system scalable, wewill develop geometric algorithms that depend on finding geometric patterns in therobot#s growth.We will show that our new robot designs and theory will work in simulation and reality within anuncontrolled environment. The results and outcomes from our research project are the following:Design and control strategies to generate emerging behaviors for robot swarms; and prototypesexperimental testbeds of multiple vehicles that operate in lab and real-world 3-D flows.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- May 15, 2023
- Source ID
- N000142312535
Entities
People
- David Saldaña
Organizations
- Lehigh University
- Office of Naval Research
- United States Navy