ASCP Team

Cyber-physical systems combine complex computing systems with physical systems, integrating sensors, actuators, advanced communication networks, and algorithms to control physical infrastructures. The CPSA team (Cyber-Physical Systems Architecture) focuses more specifically on the following aspects:

• System architecture design
• Modeling of these systems and performance evaluation
• Control of exchanges between devices using suitable and efficient protocols

The team adopts a holistic approach and addresses various aspects of cyber-physical environments by relying on multiple competencies in the following fields:

• Systems engineering

• Networking, protocols, and modeling and performance evaluation methods

• Autonomic management of cooperative and collaborative systems

Within the resulting research themes, the team’s scientific contributions are more particularly centered on:

• Protocols, modeling, and performance evaluation
• Autonomic systems and digital twins
• Applied AI for distributed systems deployed as the foundation of cyber-physical systems

Protocols, Modeling, and Performance Evaluation

Resource management and temporal constraints, the need for reliable communications, and the requirement to adapt to different deployment contexts and environments call for efficient, adaptive, and scalable protocols. The study of these systems, their components, and their interactions requires appropriate models in order to enable performance evaluation. The CPSA team uses the most suitable formalisms and tools depending on the properties of the systems under study and the targeted objectives, such as simulation, numerical analysis, or probabilistic formalisms.

Autonomic Systems and Digital Twins

Autonomic cyber-physical systems are systems capable of self-configuration and adaptation in response to dynamic changes in their environment, enabling them to meet their functional objectives.
We design and develop digital twins, based on mathematical models and/or simulation or emulation, capable of mimicking the behavior of autonomic systems. These digital twins must be able to represent heterogeneous cyber-physical systems applied to different application domains, and enable supervision and analysis of contextual data as well as the planning and execution of self-adaptation actions. These actions must meet both functional requirements and non-functional requirements such as performance, quality of service, scalability, reliability, fault tolerance, and energy consumption. Among the properties that these systems must guarantee are:

• Self-configuration and self-optimization
• Self-diagnosis and self-healing
• Self-protection
• Learning and decision-making for self-adaptation

Applied AI for Distributed Systems Deployed as the Foundation of Cyber-Physical Systems

Cyber-physical systems are systems in which data collection, analysis, management, and exploitation are key components. We focus on artificial intelligence applied to distributed systems that are deployed as the foundation of cyber-physical systems in order to:

• Design innovative methodologies for advanced management and processing of sensor data in an Internet of Things context, for applications related to agriculture, the environment, and autonomic systems;
• Develop efficient processing techniques using embedded intelligence in constrained devices (size, energy, computational power, etc.);
• More automatically identify application profiles, characteristics, and requirements to improve energy management;
• Transform digital twin models into more intelligent digital twin models with enhanced adaptation capabilities, enabling the dynamic integration of new data collected from real systems.