Cyber-Physical Systems (CPSs)  are heterogeneous  systems that present a strong interaction between the mechanical, electronic and communication components that compose them. Normally, these systems are formed by a set of devices that include sensors, actuators, processing units, control units, and communication components that support controlling and interacting with the physical process or system. Therefore, their study implies an intrinsic multidisciplinary approach, as it involves knowledge and skills from different disciplines such as computation, communications, or control engineering.

In recent years, both academia and industry are giving enormous importance to CPSs in numerous fields of application to promote the development of key technologies and applications, as well as new systems that are more intelligent, reliable, secure, and robust. However, despite the advances made so far, the control of CPSs presents numerous theoretical and technological challenges due to their inherent complexity and heterogeneity.

The main objective of the project is the design of architectures, strategies, new control laws, and communication protocols for cooperative CPSs that present communication, computing, and energy consumption restrictions with a partial knowledge of both the global system and the environment. The new strategies combine different machine learning techniques, which provide algorithms with the ability to learn and adapt, with event-based control schemes so that developments achieve feasible and robust solutions. The idea is to approach the control problem in a holistic and more generalist way, taking advantage of the potential offered by the different techniques that consider varied sources of information from the environment of the system to be controlled.

An important part of the project focuses on experimental verification in real environments in which the research group has proven experience. For this purpose, different tasks can be defined to achieve the following main objectives of this project:

1. Event-based identification and control of CPSs. From the point of view of Control Engineering, CPSs are essentially Event-Based Control Systems (EBCS). With this perspective, new dynamic identification techniques will be developed, as well as control strategies; both based on events, so that they can adapt to the changing dynamics of the system in accordance with the established specifications.
2.  Machine Learning (ML) and CPSs. Efforts will be made to integrate cooperative control techniques into CPSs using ML methods with the intention to both optimize control actions and identify and calibrate models if their dynamics are not fully known.
3. Multiagent CPSs environment. We will complete and get in production a multiagent environment formed by air drones and ground robots. This environment is currently under development and once finished, it would serve in this project to experimentally validate the results obtained in objectives 1 and 2.
4. Remote labs as CPSs. The development of a new paradigm of remote laboratories considered as CPSs with communications protocols and control strategies both based on events will be proposed. Likewise, a global architecture for the platform that would allow to remotely operate the environments established in objective 3 will be designed.