Research Projects

Cooperative control of the multi-robot system (MRS), as a class of multi-agent system (MAS) has gained immense interest in the research community due to its advantageous applications over a single-agent system due to its shortcomings like- limited computational and decision making capabilities, less coverage of the field area, overburden of the tasks and hence, high cost of the hardware. Moreover, this centralized approach is often prone to failures, compromising overall mission goals. Whereas the cooperative control majorly deploy distributed architecture and implementation through coordination and collective goal planning and execution. Thus, such decision making strategies play an important role in achieving some complex collective tasks and achieve behaviours seen in nature such as multi-agent based rendezvous, flocking, containment, formation control, attitude synchronization, target-capturing, etc. In an MRS having distributed architecture, the leader-follower approach is widely preferred as it combines some advantages of centralized scheme while keeping the distributed framework intact through clear work division rules. In a containment problem, this architecture achieves a specific task, i.e. the followers are ensured to be driven within the convex hull of multiple maneuvering leaders. This project aims to develop a robust and scalable containment control framework for a heterogeneous team of robots, operating within a ROS-based Cyber-Physical System (CPS). The objective is to ensure that followers reliably converge within the convex hull defined by multiple maneuvering leaders, even under real-world uncertainties, sensing limitations, and dynamic communication topologies etc. while maintaining scalability. Over a proposed period of three years timeline, the project will advance in four major phases, each aligned to achieve the clear scientific milestones as follows: Control Algorithm Design: Development of novel containment control strategies for heterogeneous and homogeneous MRS to guarantee convergence within known bounds. Building upon the recent successful results on binary measurement based containment control, we will integrate novel event-triggered strategies to care of sensing and communication constraints. This in turn also lessens actuator and sensor usage and thereby reducing overall bandwidth requirements. Robustness Under Adversities: Integration of discontinuous controllers, and adaptive or learning methods to counteract disturbances, actuator faults, and cyber-attacks that the real-world implementation will face. This also will incorporate the robust control formulation at agent level. Simulation and Validation: Extensive Software-in-the-Loop (SITL) simulations will be performed by using Gazebo and ROS to evaluate the developed containment control strategies' performance for both homogeneous and heterogeneous teams in dynamic environments. Real-Time Experiments: Scalable experimental validation in ROS based CPS framework, using a swarm of ground and aerial robots, tested under both kinematic and dynamic models, to validate each developed containment control strategy and finally culminating to the validation of hierarchical containment that implements multi-layered decision making and coordination among the heterogeneous MRS. By the end of the project, as mentioned above, the proposed methodologies will be rigorously validated through experiments and it is planned to realize it using up to ten heterogeneous robots for scalability checks- establishing a comprehensive foundation for deployment in real-world CPS applications spanning industrial automation, disaster response, and autonomous surveillance. The outcomes of this proposal aim to solve the real-world problems on swarm robots and contribute to the technology development in this state-of-the-art research that well aligns with "Smart cities", "Digital India", and "National Mission on Interdisciplinary Cyber-Physical Systems (NM-ICPS)" of GOI.