The latest commercially available technologies in the field of control systems, communication systems and embedded computing offered a significant stimulation on the development of autonomous systems, enabling also the research on the development of cooperative robotic systems, where the capabilities are measured in term of the team rather than a single robot.
Cooperating robotic systems are really complex and it becomes fundamental to simulate any mission stage, exploiting the benefits of simulations like repeatability, modularity and low cost. Furthermore, the design of a UAV/UGV requires a periodic review/re-design of hardware/software in order to open the systems to new attractive and challenging scenarios.
Because of the need to have a high-realistic simulation environment specifically designed for cooperating systems and a rational process to fast model and test novel features for cooperative robotics, in this project we developed a simulation framework that combines the high realism of the simulations carried out in a three-dimensional virtual environment, with the easiness of Simulink for fast prototyping of control systems.
We started with the analysis and 3D modelling of different unmanned robots, providing, for each one, a physical description in terms of dimension, mass, inertia, number of actuators and a set of input/output variables. These set of parameters allowed the physical engine (running on the simulator) to correctly estimate the dynamic behavior of each robot.
We then integrated a software module for connecting each simulated robot with the corresponding control system running on MatLab/Simulink. The set of input/output parameters, exchanged between the simulation environment and MatLab/Simulink, allowed to test several control strategies with minimum effort.
The high modularity of the simulation framework and the integration with MatLab/Simulink and third-part softwares allowed gray-box/black-box designing of algorithms for control systems.
Furthermore, it allowed to convert the control software prototyped in the simulated scenario into real applications using the tools provided by MatLab/Simulink.
We used the developed framework to successfully test some cooperative control strategies such as formation control of many UAVs using Networked Decentralized Model Predictive Control, and mission management for building inspection using Finite State Machines.
Our simulation framework for the development of cooperative robots is a flexible tool, which allows to design and validate cooperative control systems. The quality of the simulation environment has been successfully demonstrated throughout different application scenarios, particularly focused on the development of cooperating unmanned aerial vehicles.
Our simulation platform is capable not only of accurate simulation of robots in complex environments, but also makes possible to address challenges that are relevant for a variety of applications of autonomous robotics, where typical operations often include the management of complex tasks. Such application concern, for example, the use of autonomous robotics for search-and-rescue operations, building inspection, power-line inspection or industrial production.
The software of the simulation framework developed and published in the above papers has been used by the Institute of Mathematics and Computer Science, University of Latvia for their work in the development and prototyping of algorithms for cooperative robotics.
The collaboration resulted in a joined publication between the Polytechnic University of Marche and the University of Latvia, entitled Testing of cooperative tasks for Unmanned Aerial and ground platforms.