2025Activity reportProject-TeamCTRL-A
RNSR: 201421117X- Research center Inria Centre at Université Grenoble Alpes
- In partnership with:Université de Grenoble Alpes
- Team name: Control for safe Autonomic computing systems
- In collaboration with:Laboratoire d'Informatique de Grenoble (LIG)
Creation of the Project-Team: 2017 June 01
Each year, Inria research teams publish an Activity Report presenting their work and results over the reporting period. These reports follow a common structure, with some optional sections depending on the specific team. They typically begin by outlining the overall objectives and research programme, including the main research themes, goals, and methodological approaches. They also describe the application domains targeted by the team, highlighting the scientific or societal contexts in which their work is situated.
The reports then present the highlights of the year, covering major scientific achievements, software developments, or teaching contributions. When relevant, they include sections on software, platforms, and open data, detailing the tools developed and how they are shared. A substantial part is dedicated to new results, where scientific contributions are described in detail, often with subsections specifying participants and associated keywords.
Finally, the Activity Report addresses funding, contracts, partnerships, and collaborations at various levels, from industrial agreements to international cooperations. It also covers dissemination and teaching activities, such as participation in scientific events, outreach, and supervision. The document concludes with a presentation of scientific production, including major publications and those produced during the year.
Keywords
Computer Science and Digital Science
- A1.1.1. Multicore, Manycore
- A1.1.2. Hardware accelerators (GPGPU, FPGA, etc.)
- A1.1.4. High performance computing
- A1.1.5. Exascale
- A1.3.5. Cloud
- A1.3.6. Fog, Edge
- A1.6. Green Computing
- A2.1.9. Synchronous languages
- A2.2. Compilation
- A2.3.1. Embedded systems
- A2.5.1. Software Architecture & Design
- A2.5.2. Component-based Design
- A2.5.4. Software Maintenance & Evolution
- A2.6.2. Middleware
- A2.6.4. Ressource management
- A4.1.2. Hardware attacks
- A4.9. Security supervision
- A4.9.1. Intrusion detection
- A4.9.2. Alert correlation
- A4.9.3. Reaction to attacks
- A6.1.2. Stochastic Modeling
- A6.2.4. Statistical methods
- A6.4. Automatic control
- A8.2.1. Operations research
- A8.2.6. Numerical methods for optimization
- A8.9. Performance evaluation
- A9.17. Cybersecurity and AI
Other Research Topics and Application Domains
- B4.5. Energy consumption
- B4.5.1. Green computing
- B5.1. Factory of the future
- B6.1. Software industry
- B6.1.1. Software engineering
- B6.1.2. Software evolution, maintenance
- B6.4. Internet of things
- B6.5. Information systems
- B9.8. Reproducibility
1 Team members, visitors, external collaborators
Research Scientists
- Eric Rutten [Team leader, INRIA, Researcher, HDR]
- Sophie Cerf [INRIA, ISFP, from Aug 2025]
- Thierry Gautier [INRIA, Researcher, from Sep 2025, HDR]
- Clement Mommessin [INRIA, Starting Research Position]
Faculty Members
- Stephane Mocanu [Team leader, GRENOBLE INP, Associate Professor Delegation, from Dec 2025, HDR]
- Raphaël Bleuse [UGA, Associate Professor Delegation, from Sep 2025]
- Raphaël Bleuse [UGA, Associate Professor, until Aug 2025]
- Gwenaël Delaval [UGA, Associate Professor]
- Stephane Mocanu [GRENOBLE INP, Associate Professor Delegation, from Sep 2025 until Nov 2025, HDR]
- Stephane Mocanu [GRENOBLE INP, Associate Professor, until Aug 2025, HDR]
Post-Doctoral Fellow
- Jolahn Vaudey [LIG, from Oct 2025]
PhD Students
- Omayma Alla [UGA]
- Robin Chaussemy [INRIA, from Oct 2025]
- Kouds Halitim [INRIA]
- Lea Astrid Kenmogne Mekemte [UGA]
- Elian Loraux [UGA, from Oct 2025]
- Nathan Rabier [INRIA, from Nov 2025]
- Jolahn Vaudey [Grenoble-INP, until Sep 2025]
Technical Staff
- Jonathan Bleuzen [INRIA, Engineer]
Interns and Apprentices
- Mahmoud Abdo [INRIA, Intern, from Feb 2025 until Aug 2025]
- Robin Chaussemy [INRIA, Intern, from Mar 2025 until Aug 2025]
- Alexis Detroyat [INRIA, Intern, from May 2025 until Jul 2025]
- Alexis Detroyat [INRIA, Intern, from Mar 2025 until Apr 2025]
- Alexis Detroyat [INRIA, Intern, until Feb 2025]
- Mohamed Abdeldjalil Maziz [INRIA, Intern, from Feb 2025 until Aug 2025]
Administrative Assistants
- Marie-Anne Dauphin-Rizzi [INRIA]
- Maria Immaculada Presseguer [INRIA]
External Collaborator
- Bogdan Robu [UGA]
2 Overall objectives
Objective: control support for autonomic computing
Ctrl-A is motivated by the observation that computing systems, large (data centers) or small (embedded), are more and more required to be adaptive to the dynamical fluctuations of their environments and workloads, evolutions of their computing infrastructures (mobile, shared, or subject to faults), or changes in application modes and functionalities. Their administration, traditionally managed by human system administrators, needs to be automated in order to be efficient, safe and responsive. Autonomic Computing 40 is the approach that emerged in the early 2000's in distributed systems to answer that challenge, in the form of feedback loops for self-administration control. These loops address objectives like self-configuration (e.g. in service-oriented systems), self-optimization (resource consumption management e.g., energy), self-healing (fault-tolerance, resilience), self-protection (security and privacy).
Therefore, there is a pressing and increasing demand for methods and tools to design controllers for self-adaptive computing systems, that ensure quality and safety of the behavior of the controlled system. The critical importance of the quality of control on performance and safety in automated systems, in computing as elsewhere, calls for a departure from traditional approaches relying on ad hoc techniques, often empirical, unsafe and application-specific solutions.
The main objective of the Ctrl-A project-team is to develop a novel framework for model-based design of controllers in Autonomic Computing, exploiting techniques from Control Theory 37, particularly Discrete Event Systems 46, but also other forms. We want to contribute generic Software Engineering methods and tools for developers to design appropriate controllers for their particular reconfigurable architectures, software or hardware, and integrate them at middleware level. We want to improve concrete usability of techniques from Control Theory by specialists of computing systems 11, and to provide tool support for our methods in the form of specification languages and compilers, as well as software architectures.
We address policies for self-configuration, self-optimization (resource management, low power), self-healing (fault tolerance) and self-protection (security).
3 Research program
Modeling and control techniques for autonomic computing
Our research activity is mainly targeted at models and architectures, with also a notable part devoted to applications and case studies, in co-operations with specialists of the application domains, either academic researchers or industrial partners (e.g., CEA, Orange labs, RTE, Qarnot Computing, Naval Group). We adopt a strategy of parallel investigation of, on the one hand, generic models and tools for the design support for control in Autonomic Computing, and, on the other hand, experimental identification of needs and validation of proposals. Therefore we have activities related to several application domains, for each of which we build co-operations with specialists, for example middleware platforms for Cloud systems 3, HPC architectures (e.g., multi-core 15), Dynamic Partial Reconfiguration in FPGA-based hardware 9 and the IoT and smart environments 47 .
The main objective of Ctrl-A translates into a number of scientific axes :
- Design support for Control in Autonomic Computing : under the angle of Models and control (e.g., Discrete Event Systems and reactive languages), or at the level of Software Components and Architectures (e.g., for separation of concerns, coordination of multiple autonomic managers : Control, ML, RJMS, or application/infrastructure-levels);
- Self-adaptative distributed systems and Cloud-Edge/HPC : e.g., RJMS-level dynamical resource harvesting in HPC clusters ; node-level energy management through RAPL ; reproducibility of experimental validation.
- CyberSecurity & Self-protection in Industrial Control Systems : intrusion detection ; automated risk analysis ; validation of conformity to IEC 62443 standard; self-protection, resilience and reaction by self-reconfiguration ; applications to Smart-Grid infrastructures ; experimental lab.
Achieving the goals of Ctrl-A requires multidisciplinarity and expertise from several domains. The expertise in Autonomic Computing and programming languages is covered internally by members of the Ctrl-A team. On the side of theoretical aspects of control, we have active external collaborations with researchers specialized in Control Theory, in the domain of Discrete Event Systems as well as in classical, continuous control. Additionally, an important requirement for our research to have impact is to have access to concrete, real-world computing systems requiring reconfiguration control. We target autonomic computing at different scales, in embedded systems or in cloud infrastructures, which are traditionally different domains. This is addressed by external collaborations, with experts in either hardware or software platforms, who are generally missing our competences on model-based control of reconfigurations.
4 Application domains
We tackle the problem of designing well-regulated and efficient self-adaptive computing systems by the development of novel strategies for their runtime management. Therefore the kind of application domains that we typically target involve computing systems with relatively coarse-grain computation tasks (e.g. image processing or HPC tasks, components or services, control functions in Industrial Control Systems). They must be run on distributed heterogeneous architectures. Runtime, unpredictable variations can come from the environment (e.g., data values, user inputs, physical sensors), the application (e.g., functional modes depending on algorithm progress, computation phases, or business processes), or the infrastructure (e.g., resource overload, faults, temperature variations, communication network variations, cyber-attacks).
The general control problem then consists of deciding at runtime the choice of which implementation or version of tasks to dynamically deploy or redeploy on which computing resources, in order to enforce high-level strategies involving objectives in terms of constraints, optimization, logical invariance or reachability. The design of such controllers involves the design of appropriate sensors and actuators in the computing infrastructures. It is based on the use of modeling and decision formalisms of different kinds according to the application characteristics.
The objectives of Ctrl-A are achieved and evaluated in both of our main application domains, thereby exhibiting their similarities from the point of view of reconfiguration control.
Self-adaptive and reconfigurable computing systems, in Cloud-Edge and HPC
One main application domain for the research of Ctrl-A concerns Cloud-Edge and High-Performance Computing. In these contexts, tasks can be achieved following a choice of implementations or versions, such as in, e.g., service oriented approaches. Each implementation has its own characteristics and requirements, e.g., w.r.t. resources consumed and QoS offered. The systems execution infrastructures present heterogeneity, with different computing processors, a variety of peripheral devices (e.g., I/O, video port, accelerators), and different means of communications. This hardware or middleware level also presents adaptation potential, e.g., in varying quantities of resources or sleep and stand-by modes.
The kinds of control problems encountered in these self-adaptive systems concern the navigation in the configurations space defined by choice points at the levels of applications, tasks, and architecture. The pace of control is more sporadic, and slower than the instruction-level computation performance inside the large-grain tasks.
In this application area, we currently focus especially on the runtime management of resources for energy objectives and digital soberness, e.g. at the level of a data-center by dynamically harvesting unused resources, or at node level by dynamically adjusting frequency under QoS constraints. Ongoing or recent cooperations in the application domain feature Qarnot Computing (challenge Inria PULSE), Orange labs, Nokia, Argonne National Laboratories (USA) (JLESC).
Cybersecurity of Industrial Control Systems,
We are developing applications in the field of cybersecurity of industrial control systems mainly in the field of intrusion detection systems (IDS), reaction to attacks and experimental lab. We are working both on manufacturing control systems like SCADA or DCS (Distributed control systems) and electrical substation protection systems.
In the manufacturing systems intrusion detection field, we worked mainly on the network level detection of stealth process-oriented attacks i.e. attacks that do not violate the syntax or semantics of communication protocols. Such attacks are sending legitimate controls in a wrong process context. We developed detection frameworks for this type of attacks by runtime-monitoring technique (PhD Oualid Koucham DGA , PhD Estelle Hotellier Naval Group) and AI techniques (ongoing Léa Kenmogne PhD – PERP Cybersecurity Superviz). Recently we started a study on host-based IDS (ongoing Omayma Alla PhD). We also study the intrusion detection methodologies for electrical substations compliant with IEC 61850 standard (PhD Maëlle Kabir-Querrec) and the ongoing Jolahn Vaudey postdoc.
On the reaction side we study the reconfiguration of industrial control systems under attack. We developed a methodology compliant with IEC 62443 standard (PhD Jolahn Vaudey) and we also worked on the dynamic reconfiguration and resource allocation of virtualized electrical substations (Postdoc Salim Chehida). Recently we started a PhD study dedicated OT (Operational Technologies) honeypot (Elian Loraux).
We also work on the risk assessment automated risk assessment and industrial systems (PhD Da Silva) and we have an important activity on the development of experimental labs (G-ICS lab).
Ongoing or recent cooperations in the application domain feature DGA, Naval Group, CEA, and RTE (the French energy transportion company).
5 Social and environmental responsibility
5.1 Footprint of research activities
In the year 2025, we are still trying to moderate the travels of the team. We favor submissions and publications in journals.
Our activities in energy-efficient management of computing infrastructures involve running experiments on large computing infrastructures such as Grid'5000 and GRICAD. We approximately spent 27,000 (Grid'5000) and 20,000 (GRICAD) core·hours of computing.
In order to diminish or limit this cost, we are working towards building up simulation techniques appropriately integrating self-adaptation and control loops, as well as environmental variability.
5.2 Impact of research results
We have research activities w.r.t. energy efficiency in computing systems, at the levels of nodes (RAPL) as well as at the higher level of computing infrastructures (RJMS, CiGri). These works are contributing to a better mastered energy consumption in computing.
On a longer term, we orient our research towards topics explicitely targeting environmental as well as social impacts, in the form of user involvement through usage choices. In line with our topic of autonomic management, self-adaptive systems and their control, for example, we consider control objectives involving trade-offs between performance or QoS and economy of resources and impact, so that users can choose a level of sobriety, and possibly limited or degraded quality, thereby allowing for potential resource and energy savings. Our ongoing cooperation with Qarnot Computing has a potential for involving not only technical considerations but also societal and regulatory constraints, or user and customer choices.
The perspectives involve the notion of computing within limits, especially when they are varying dynamically, and which can be undergone (e.g., resilience when submitted to cyber-attacks or faults) or chosen (e.g., accepting lower quality outside of phases requiring higher levels due to urgency). Our cooperation with RTE (French electricity transportation industrial), in the framework of the Tasting project of the PEPR TASE, is exploring the impact of the notion of consumption flexibility (required from consuming industries) on the management of computations in Data Centers, which is a rapidly growing consuming industry, especially in Ile-de-France.
6 Highlights of the year
6.1 Awards
The paper “Curricula Developments regarding Industry 4.0 in an Asian-European partnership in 3 countries in South-East Asia” by Chaisricharoen and al. (co-authored by Stéphane Mocanu ) 21 obtained the best paper award of the EAEEIE 2025 conference.
6.2 PhD defenses
Jolahn Vaudey defended his PhD “Reconfiguration of industrial systems as a reaction to cyberattacks” (theses.fr/s360131) on October 14th 2025.
7 Latest software developments, platforms, open data
We continue to maintain and develop our software packages. The only new addition in 2025 is the automatic PLC intrumentation code.
7.1 Latest software developments
7.1.1 Heptagon
-
Keywords:
Compilers, Synchronous Language, Controller synthesis
-
Functional Description:
Heptagon is an experimental language for the implementation of embedded real-time reactive systems. It is developed inside the Synchronics large-scale initiative, in collaboration with Inria Rhones-Alpes. It is essentially a subset of Lucid Synchrone, without type inference, type polymorphism and higher-order. It is thus a Lustre-like language extended with hierchical automata in a form very close to SCADE 6. The intention for making this new language and compiler is to develop new aggressive optimization techniques for sequential C code and compilation methods for generating parallel code for different platforms. This explains much of the simplifications we have made in order to ease the development of compilation techniques.
The current version of the compiler includes the following features: - Inclusion of discrete controller synthesis within the compilation: the language is equipped with a behavioral contract mechanisms, where assumptions can be described, as well as an "enforce" property part. The semantics of this latter is that the property should be enforced by controlling the behaviour of the node equipped with the contract. This property will be enforced by an automatically built controller, which will act on free controllable variables given by the programmer. This extension has been named BZR in previous works. - Expression and compilation of array values with modular memory optimization. The language allows the expression and operations on arrays (access, modification, iterators). With the use of location annotations, the programmer can avoid unnecessary array copies.
- URL:
-
Contact:
Gwenaël Delaval
-
Participants:
Gwenaël Delaval, Marc Pouzet, 5 anonymous participants
-
Partners:
UGA, ENS Paris, Inria, LIG
7.1.2 DSL62443
-
Name:
DSL for modeling IEC 62443 compliant ICS instances, in a zone/conduit model
-
Keyword:
DSL
-
Functional Description:
Source code for an eclipse plugin, allowing the user to describe ICS instances that follows the IEC 62443 zone/conduit model. Uses xtext for the textual IDE, and Sirius for the graphical one. Also contains an application, using constraint programming to reconfigure a modeled system under attack. This controller is generated from a description.
- URL:
-
Contact:
Jolahn Vaudey
7.1.3 RTUShark
-
Name:
Modbus RTU network trafic capture
-
Keyword:
Network monitoring
-
Functional Description:
Allow to capture Modbus RTU network trafic, and save it in .pcap format, using a USB-RS485 adapter.
- URL:
-
Contact:
Stephane Mocanu
7.1.4 GICS_HIL
-
Name:
Hardware in the loop simulation system
-
Keyword:
Hardware and Software Platform
-
Functional Description:
We offer a set of electronic boards (electronic schematics and Gerber files for PCB manufacturing) that enable the emulation of digital and analog sensors and actuators, thus allowing interfacing between industrial control/command/protection equipment and physical process simulation software. The system core is an electronic card inspired by the STM32_Discovery evaluation board. The embedded firmware can be configured to manage I/O from industrial programmable logic controllers (24V digital or +/-10V analog) or three-phase current and voltage sensors compatible with electrical substation protection relays. Two power cards allow the signals to be adapted to current/voltage transformer standards: 0..1 A and +/- 48V (48V instead of 110V for compatibility with electrical safety standards for university workbench). The cards interface with the simulation software via UDP communication. Interfacing software with Factory I/O and Home I/O is provided, as well as software for controlling three-phase signals. Interfacing with Matlab/Simulink is direct via Simulink UDP blocks. Interfacing with Modelica is also direct via the UDPSend and UDP Receive blocks of the DeviceDrivers.Communication package
- URL:
- Publications:
-
Contact:
Stephane Mocanu
7.1.5 PLCOPENXML instrumentation tools
-
Name:
Python scripts, instrumenting CEI 61131-10 PLCOPENXML files to allow for context migration.
-
Keyword:
Instrumentation
-
Functional Description:
Python scripts, that take as input a PLCOPENXML description of a PLC project. Then, for each application in the project: - Asks the user to determine its limits (SFC programs and relevant variables) - Instrument the program to store the execution context in an accessible area (here, through Modbus) - Instrument the program to allow migration - Store these information in a serialized file, as a runnable snapshot manager. This manager can a) regularly take snapshots of the corresponding application's state and b) migrate this execution context to another PLC.
- URL:
- Publication:
-
Contact:
Jolahn Vaudey
7.2 New platforms
7.2.1 G-ICS: GreEn-ER Industrial Control systems Sandbox
Participants: Stéphane Mocanu.
GreEn-ER Industrial Control Systems Sandbox (G-ICS) is a teaching and research platform that brings together a hundred control, command, and industrial supervision devices of multi-protocol and multi-manufacturer that can be flexibly coupled with a hardware-in-the-loop simulation system. The software simulation of processes can be carried out with commercial simulators (Matlab/Simulink, Dymola) or open-source (Scilab or Modelica) as well as with virtualization environments (Factory I/O or Home I/O). Thus, system architectures covering industrial fields ranging from home automation and electrical distribution in buildings to manufacturing industry and smart grids can be realized.
Hardware Environment.
The platform consists of several distinct models. The following sections detail the main use cases. In addition to the described models, a significant amount of automation equipment is available for custom experiment realization.
2D Robot Models
Three Schneider MD1ADAX2M 2 Axis Machines educational models including two industrial axes, an M340 PLC, two speed drives, and an HMI. Implemented protocols: CANOpen and Modbus/TCP.
The image depicts a detailed view of an industrial control panel and a machine setup. The control panel contains various electrical and electronic components such as switches, circuit breakers, and controllers. Labels on the image identify parts like the M340 automation unit, Ethernet switch, power supplies, differential circuit breaker, and safety modules. The machine setup includes features such as a multi-axis system, safety interrupters, a transparent polycarbonate door, and a marking unit. The control panel and machine components are interconnected with wires and cables, and the setup appears to be for an industrial automation or manufacturing process. (Description generated at January 18th, 2026 by Albert AI with the model Mistral-Small-3.2-24B)
FischerTechnik Models
Scaled-down physical processes: Six FischerTechnik factory simulator cases1. Sensors/actuators interfaced with TM3BCEIP remote I/O units (Figure 2). Communication based on Modbus/TCP.
The image displays a interconnected PLC network centered around a FischerTechnik Industry 4.0 factory simulation. Sensors and actuators of each subprocess are acquired by RTU then read by the four PLC that control the process. An HMI allows to monitor the process state. Each supprocess is identified by a color.
Hardware-in-the-Loop Models
About fifty programmable logic controllers (PLCs) from Schneider, Siemens, Wago, and ABB connected to electronic interfaces allowing coupling with process simulators. Home I/O, Factory I/O simulator licenses available or interfacing with Matlab/Simulink, Modelica, etc. The principle is presented in Figure 3.
The image depicts a centralized automation system with various interconnected components. At the center there are three control units linked to electronic interfacing cards and to multiple devices and interfaces, including databases, personal computers, HMIs. Under the interfacing cards are logos and icons representing programming languages and software tools, such as MATLAB Simulink, Python, C++, and Modelica, highlighting the interfacing of industrial controllers with of different programning technologies via the electronic cards.
Multiprotocol Models
Multiprotocol models: an internally made "SCADA suitcase" model is available (10 copies - Figure 4). The model allows studying and carrying out attacks on hierarchical and distributed multi-protocol architectures (Modbus TCP, CANOpen, OPC UA, Modbus RTU).
The image depicts an industrial supervision and control system. It features a local Modbus/TCP and OPC UA network connecting a PC supervision screen to a network switch. The switch connects to two automation devices, API M340 and API M251, via an HMI (Human-Machine Interface). These devices link to sensors/actuators through Modbus/RTU and CanOpen field buses, controlling a piloted process. Connections include both digital (TOR) and analog signals. (Description generated at January 22nd, 2026 by Albert AI with the model Mistral-Small-3.2-24B)
Smart-grid Platform IEC 61850
About 30 protection relays and current and voltage measurement units are available for creating models of protection and control infrastructures for electrical networks based on IEC 61850 communication protocols and redundant HSR and PRP networks. One of the models constructed from several standalone measurement units (SAMU) and protection relays (IED) is presented in Figure 5. The redundant network is accessible for frame capture and attack deployment via a Redundancy Box (RedBox).
The image shows a network setup involving industrial devices. On the left is a diagram with SCADA, PRP (Parallel Redundancy Protocol), and GOOSE (Generic Object Oriented Substation Event) connections. Key components include IEDs (Intelligent Electronic Devices), SAMUs (Synchronized Measurement Units), and RedBoxes. On the right is a rack with actual equipment labeled RedBox, IED, and SAMU, connected by numerous wires. The RedBox serves as a network device, while IEDs and SAMUs are connected to monitor and control electrical systems. (Description generated at January 18th, 2026 by Albert AI with the model Mistral-Small-3.2-24B) (Description generated at January 18th, 2026 by Albert AI with the model Mistral-Small-3.2-24B)
Safety Automation Tesbed
Three complete testbeds (sensors, actuators, and programmable controllers) from PILZ brand as well as several Schneider and Siemens safety PLCs are available for demonstrating attacks on safety command functions (Figure 6).
The image shows a training workbench dedicated to safety related control systems. Are visible : a safety Pilz PLC, an HMI, a security badge reader with badges and various sefety sensort : an emergency stop, a safety gate, and an enabling switch.
Software Environment
- Control/command component programming environments (Schneider, Siemens, Pilz, ABB, Kepware)
- Industrial supervision software: PCVues, WinCC
- OPC servers Kepware and Matrikon
Monitoring and Development Tools
Wireshark available on all computers with a local partial UMAS dissector, admin access to switches and the ability to configure additional ports in monitoring mode. Snort and zeek instances.
Usage Constraints
The platform is connected to Grenoble-INP's classroom PCs. Although it is possible to temporarily disconnect the PCs, for safety reasons, the deployment of ransomware and auto-replicating viruses will not be possible. DoS attacks are possible under certain conditions: the classroom computers being connected to the model networks and to the Grenoble-INP network, to avoid disruptions to the institution's network it is necessary, beforehand, to physically disconnect the classroom computers. Dedicated computers (not managed by the institution) must be used for DoS type attacks to prevent alerting Grenoble-INP's EDR. In order to ensure the safety of the equipment we are not allowing destructive attacks (loading corrupted firmwares).
Description of Normal Use Cases
For industrial systems, normal activity corresponds to the normal operation of the physical process. Traffic corresponds to the communication between industrial controllers or with sensors/actuators or even with industrial supervision.
Description of Implemented Attacks
The implemented attacks are forcings of input values (sensors), output values (actuators), or internal variables of the controllers. The currently supported network protocols are Modbus/TCP, UMAS (partially), CANOpen, GOOSE, and SV (version 61850-9-2LE).
Modbus/TCP Attacks
This type of attack targets the communication network between industrial supervision and programmable PLCs and uses writing to the internal variables of the PLCs via Modbus/TCP clients (free software on the Internet or Metasploit). The attacks use "legitimate" commands (e.g., opening a valve or stopping a motor) in contexts where the action can damage the physical process. Concrete examples of attacks on Modbus/TCP are detailed in 41, 43, 42, 38.
CANOpen Attacks
This type of attack directly targets the traffic between programmable PLCs and local loop controllers. It is currently implemented only on the "2D Robots" model. The attack consists of injecting commands on the CAN bus that, for example, cause a speed drive to fail or change its operating mode, modify or stop ongoing movements, perform forbidden movements or movements at too high speeds, inject false sensor values, or desynchronize the movements of the two axes. A total of 17 types of attacks are implemented and described in 38.
GOOSE Attacks
The GOOSE protocol is an Ethernet multicast protocol used for transmitting events in an IEC 61850 electrical substation. Essentially, it involves triggering circuit breakers (trip). The GOOSE frames are identified using two counters (one for the frames and another for the state changes-events). The attack consists of usurping the legitimate sequence of GOOSE frames by injecting frames that cause a circuit breaker to trip (a single frame is sufficient). The implementation and detection have been presented in 39.
SMV Attacks
The SMV protocol is an Ethernet multicast used for transmitting samples of current and voltage measurements in IEC 61850 substations. The SMV frames are identified by counters. Two types of attacks are available: injecting false measurements and Ethernet flooding. Injecting false measurements consists of usurping the legitimate flow of SMV and, therefore, injecting a flow with false measurements. Ethernet flooding targets high availability networks HSR and PRP. In particular, "double ring" HSR networks are very vulnerable to Ethernet floods. The attacks and their effects on electrical protections and network infrastructures have been presented in 44 and 45.
Several public datasets where generated on the platform.
Modbus Datasets
Datasets corresponding to the work of Oualid Koucham's thesis2 are available in pcap format. The datasets and associated attacks are detailed in section 3.3.3 (Implementation and datasets) of the thesis. The datasets were generated on a Hardware-in-the-Loop model also described in the thesis with Modbus/TCP attacks. The model is built on the principle illustrated in the referenced figure coupled with a software simulation of a simplified version of a well-known physical process in control engineering: the Tennessee-Eastman distillation column. A reduced demo of the attacks and detection is available in virtual machine format (replay of the datasets, deliverable of the ANR ASTRID SACADE project).
CANOpen Datasets
Datasets of ModbusTCP and CANOpen attacks generated on the 3D Robots model come from the thesis work of Estelle Hôtellier. The datasets and a detailed description of the attacks are available on request.
IEC 61850 Datasets
A dataset corresponding to a type of attack involving the injection of false trip signals in a GOOSE flow comes from the thesis of Maëlle Kebir-Querrec3.
Several datasets are available for attacks involving the injection of false sensor values in SMV flows (IEC 61850) as well as Ethernet flood type attacks in high reliability HSR/PRP networks.
7.3 Open data
IDS calibration and benchmark ICS dataset
-
Contributors:
Stephane Mocanu, Léa Kenmogne, Oualid Koucham
-
Description:
Datasets containing industrial traffic captured in a simple industrial system. Two datasets are provided : normal traffic dataset for training of the Intrusion Detection System and the dataset containing the process oriented attacks (actuators manipulation) for IDS banchmarking. Both datasets are provided in two versions : the original network trafic capture (pcap file) and the extracted features (sensor and actuators values) CVS files.
- Dataset PID (DOI,...):
- Publications:
GICS Intrusion Detection Datasets
-
Contributors:
Oualid Koucham, Stephane Mocanu
-
Description:
These datasets include network traces collected at the ENSE3 GICS platform for the purposes of evaluating an intrusion detection system (IDS) for ICS. The network traces capture the behavior of an ICS test bed under attacks targeting the physical process. The test bed is implemented in GICS and is comprised of several controllers (Schneider M340/M580, Wago IPC-C6, Siemens, etc.) along with supervisory machines, engineering workstations and human machine interfaces (HMIs). Each controller sends commands and receives sensor information, via I/O interface cards, from a real-time OpenModelica simulation of a complex physical process representing a complex chemical plant. The traces contain, among other protocols, Modbus traffic carrying attacks violating the specifications of the underlying physical process. This is performed by sending a sequence of Modbus commands from workstations to controllers running the control logics which steer the process. Two types of attacks are contained in these datasets. The first type of attacks violates qualitative temporal constraints on the behavior of the physical process. Examples of such attacks include opening simultaneously two valves or stopping a motor before its due time. The second type of attacks violates quantitative temporal constraints. For example, the traces include attacks that wear a valve by quickly opening and closing it. The contents of the datasets is as follows: One capture free from attacks and containing only legitimate traffic and Four captures containing attacks.
- Dataset PID (DOI,...):
- Publications:
8 New results
8.1 Design support for Control in Autonomic Computing
8.1.1 Autonomic Resource Harvesting in HPC: Control Methods and their Reusability
Participants: Raphaël Bleuse, Sophie Cerf, Quentin Guilloteau, Rosa Pagano, Bogdan Robu, Eric Rutten.
Using control theory methods to adapt to unpredictable variations in resource managment allows for the design of well-founded autonomic managers. Choosing the relevant approach is daunting due to the variety of existing controllers. The criteria are of different natures, involving performance and efficiency, but also required expertise in control theory, and reusability or portability between sub-systems. We study how reusability relates to the adaptivity and robustness properties in control. We compare various control designs:
- the classic Proportional-Integral-Derivative (PID) control,
- its upgrade as adaptive control,
- and Model-Free Control (MFC).
We perform experimental evaluation and compare performance and reusability. Trade-offs are found on different criteria: while adaptive control is largely portable, its design complexity is significant for non-experts; PID control has good nominal performance, yet its portability is limited; MFC requires few competences to be used, but cannot provide strong guarantees.
This comparative study has been published in 17.
8.2 Self-adaptative distributed systems and Cloud-Edge/HPC
8.2.1 Mitigation of I/O Congestion Using Control Theory
Participants: Raphaël Bleuse, Sophie Cerf, Thomas Collignon, Kouds Halitim, Bogdan Robu, Eric Rutten, Lionel Seinturier, Alexandre van Kempen.
Efficient data access in High-Performance Computing (HPC) systems is essential to the performance of intensive computing tasks. Traditional optimizations of the I/O stack aim to improve peak performance but are often workload specific and require deep expertise, making them difficult to generalize. In shared HPC environments, resource congestion can lead to unpredictable performance, causing slowdowns and timeouts. To address these challenges, a self-adaptive approach based on Control Theory is proposed to dynamically regulate client-side I/O rates. Our approach leverages a small set of runtime system load metrics to reduce congestion and enhance performance stability.
We study this problem in two distinct experimental setups:
- A single-node system, as a preliminary work 26, 33.
- A multi-node cluster, where we evaluate a representative workload on a real testbed. Experimental results demonstrate that our method effectively mitigates I/O congestion, reducing total runtime by up to 20% and lowering tail latency, while maintaining stable performance. These results have been published in 22.
8.2.2 Graceful application degradation in HPC with Power Capping
Participants: Raphaël Bleuse, Sophie Cerf, Kouds Halitim, Mohamed Abdeldjalil Maziz, Bogdan Robu, Eric Rutten.
We study control-based approaches that regulate the power usage of an HPC application to achieve a graceful degradation of its performance, i.e. allowing significant energy savings with limited increase in execution time.
Power capping is performed using the RAPL (Running Average Power Limit) actuation and sensing mechanism, which is known to have noise and inaccuracies. We address the challenges of noise and actuator inaccuracies with an approach that incorporates a cascade control design, allowing to have an inner loop specially designed to ensure that the actuator effisciently reach the desired action value in a short time (compared to the main regulation loop).
Additionaly, the number of processors in an HPC setup can vary, which implies that the control has to handle a variable number of power signals (i.e. output) for a same input. The control design results in a robust cascade control approach. The system can be divided into two subsystems: the RAPL actuator and the HPC plant. The first subsystem is a Single-Input Multiple-Output (SIMO) system due to the different power values collected from power sensors, while the second subsystem is a Multiple-Input Single-Output (MISO) system between the power measures and the application's progress.
This approach of robust cascade control for variable dimension systems has been published in 23.
Previous works study the power control on various benchmarks, i.e. either compute-intensive or memory-intensive. We tackled a more realistic scenario by considering a workload that vary in computational intensity, alternating between compute-bound and memory-bound phases. Two adequate control strategies are explored: a gain-scheduled PI controller and a polytopic LPV controller, both modulated by a scheduling parameter, which represents the ratio of compute to memory activity. This was the work done by Mohamed Abdeldjalil Maziz during his internship 34.
8.2.3 Stochastic Control for Resource Harvesting
Participants: Mahmoud Abdo, Raphaël Bleuse, Sophie Cerf, Kouds Halitim, Bogdan Robu, Eric Rutten.
This works builds upon previous works on the Cigri/OAR composition to harvest idle resources in HPC clusters. One promising approach is to fill idle resources by injecting small, flexible, independent, and interruptible jobs that have no strict time constraints. However, managing the injection of these jobs is challenging due to the uncertain nature of job characteristics, such as execution times and resource consumption. Additionally, process noise-resulting from system complexity and the arrival and execution of varying external workloads-can interrupt or terminate these filler jobs.
We propose and evaluate, using real data, a Stochastic Model Predictive Control (SMPC) approach that addresses system uncertainty and incorporates a feed-forward compensation mechanism for disturbance rejection. The proposed algorithm shows promising results: it ensures a platform usage rate of 98%, significantly improving overall resource efficiency and reducing the number of early terminated jobs compared to previous work.
The results were published in 27. The formulation of the Model Predictive Controller has been studied by Mahmoud Abdo during his internship 30.
8.2.4 Modeling and Simulation of HPC Jobs Arrival
Participants: Raphaël Bleuse, Robin Chaussemy, Franck Corset.
The contributions of this work are twofold. First we consolidate in a single database the exploitation traces of the GRICAD clusters for the past ten years. The consolidation process merges the exploitation databases of each cluster's OAR database and the exploitation database of the job injector Cigri. Second, we model the arrival of jobs on the platform as a homogeneous Poisson process. We then simulate jobs' arrival with a marked Poisson process to account for the polling frequency of the RJMS.
Preliminary results are available the Master thesis of Robin Chaussemy31. The modeling lays the foundation for refining control signals.
8.2.5 Flexibility in Data-Centers for Digital Soberness
Participants: Raphaël Bleuse, Alexis Detroyat, Clement Mommessin.
We collaborate with RTE (Réseau de transport d'électricité) to study the flexibility capabilities of data-centers. One of the aspect we studied is the adaptation of the computation load to a time-varying power budget. Alexis Detroyat worked on the characterization of the relationship between the power cap applied to a server and the runtime of jobs executed on that server. This characterization has been conducted on a diverse set of hardware from the Grid'5000 testbed using workloads from the NAS Parallel Benchmark. This work allows us to identify, with respect to the hardware, a range of applicable power cap at a reasonable increase of the execution time of the jobs. These experiments are detailed in the internship report 32.
Such experiments are a step towards defining ranges of flexibility to improve or design new power-aware scheduling algorithms for computing platforms.
8.2.6 Performance portable batched linear algebra kernels for transport sweeps using Kokkos
Participants: Thierry Gautier.
The paper 28 describes the development of performance portable batched linear algebra kernels for SN-DG neutron transport sweeps using Kokkos. We establish a new sweep algorithm for GPUs that relies on batched linear algebra kernels. We implement an optimized batched gesv solver for small linear systems that builds upon state-of-the-art algorithms. Our implementation achieves high performance by minimizing global memory traffic and maximizing the amount of computations done at compile-time. We assess the performance of the batched gesv kernel on NVIDIA and AMD GPUs. We show that our custom implementation outperforms state-of-the-art linear algebra libraries on these architectures. The performance of the new GPU sweep implementation is assessed on the H100 and MI300A GPUs. We demonstrate that it is able to achieve high performance on both architectures, and is competitive with an optimized multithreaded CPU implementation on a 128-core AMD Genoa CPU node.
This work, published in a SC'25 workshop last october, is the subject of PhD Gabiel SUAU (CEA) supervised by Thierry Gautier and it has begun in the Avalon project team.
8.3 Cybersecurity of Industrial Control Systems
8.3.1 Safety-Security Convergence: Automation of IEC 62443-3-2
Participants: Da Silva Mike, Stephane Mocanu, Maxime Puys, Thevenon Pierre-Henri.
We develop a method to automate industrial cybersecurity risk assessment as specified in the IEC 62443-3-2 standard, which is widely used in the industrial cybersecurity domain. By automating parts of these risk assessment processes, we can reduce the error-prone manual efforts and increase the consistency of risk assessment. More specifically, the proposed risk assessment comprises three parts which, respectively:
- identify the specific vulnerabilities of industrial control systems,
- determine the attack scenarios that compromise the safety of the system
- assess whether the attack scenarios are tolerable by the organization's policy.
In the first part, we automated the entire threat modeling process of file called PLCOpen. This automation of the Microsoft Threat Modeling Tool process Threat Modeling Tool by developing an automatable method for building the system model, in the form of a data flow diagram, from a standard XML file called PLCOpen. This automation of the Microsoft Threat Modeling Tool process enables us to automate vulnerability identification for industrial control systems.
In the second part, we enhance a previous work that generates theoretical safety-compromising attack scenarios by building a complete attack scenario from system vulnerabilities to safety compromise.
Finally, in the third part, we rank the attack scenarios using a specific risk matrix in order to determine which scenarios exceed the risk tolerable by the organization and therefore require additional controls.
The results were published in 16
8.3.2 Host-based Intrusion Detection for Industrial Control Systems
Participants: Omayma Alla, Stephane Mocanu.
Traditionally, intrusion detection in Industrial control systems is performed at network level due to the fact that the embedded OS of the Programmable Logic Controllers (PLCs) is accessible only to the manufacturers.
We start evaluating the posibility to effectively incorporating lightweight Host-Based Intrusion Detection Systems (HIDS) into PLCs. This technique provides a more comprehensive and rapid degree of security through monitoring device activity in real time. The methodology uses formal specification methods to generate attributes from industrial control logic (in particular written in SFC language) and regulations that can be verified at runtime.
The overall idea was published in 20. A first IDS was implemented as a PLC separate task and a proof of concept was implemented on the G-ICS testbed. The result was accepted for publication in January 2026 at the ICIT26 conference.
8.3.3 An Explainable Approach to Process-Oriented Attacks in Industrial Control Systems using SHAP and LIME
Participants: Lea Astrid Kenmogne Mekemte, Stephane Mocanu.
Process oriented attacks are attacks that target specifically the controlled physical process in an industrial control system. Typically they are “stealth” attacks in the sense that they do not violate the syntax or the semantics of communication protocols. They are sending legal frames (controls to actuators) but in an inapropriate context. In previous approaches (42, 38) we use runtime-monitoring of security specification patterns extracted either from devices documentation and standards or mining of normal execution traces. The problem of these approaches is that, in general, the full set of security patterns cannot be obtained. Specification mining will produce a lot of false positive or inconsitent patterns while extraction from device documentation and standars is manual and limited. On the other hand control programs specifications are rarely available and reverse engineering of the specifications from PLC programs is very costly and cannot be automatized.
We propose an approach to detect process oriented attacks in industrial control systems based on explainable artificial intelligence (XAI) using LIME and SHAP methods to understand the model’s predictions. The experimental evaluation, conducted on a simulated industrial process, demonstrates that our approach provides good detection accuracy and a comprehensive explanation of attacks being able to rebuild a large number of security pattern from the explaination of the alerts.
The methodology and experimental validation were presented in 24.
8.3.4 Reconfiguration of Firewall Filter Rules as a Response to Industrial Control System Intrusion
Participants: Jolahn Vaudey, Stephane Mocanu, Gwenaël Delaval, Stephane Mocanu.
As the IEC 62443 becomes the “de facto” information security standard in industrial control systems a need of supporting tools and methodologies for the implementation of the standard requirements has arised.
This result adresses the network segmentation (called “Restricted Data flow”) requirements of the standard in particular reacting to the attack. Starting with a formal description of the IEC 62443 compliant ICS under study, we automate the creation of all necessary filter rules, and the adaptation to application migration and compromised device isolation. This approach is tested on a small scale installation supervising a physical process with industrial hardware.
The methodology and experimental validation were presented in 25.
8.3.5 Self-Reconfiguration of Industrial Control Systems as a Response to Cyberattacks
Participants: Jolahn Vaudey, Stephane Mocanu, Gwenaël Delaval, Eric Rutten.
In the same vein of system reconfiguration in case of an attack as the previous result, we propose a system that, upon detection of a compromised component, dynamically reconfigures itself to maintain functionality. Our approach leverages the increasing virtualization of ICS to migrate tasks from compromised devices to healthy ones, ensuring continued operation while containing the attack. We model the reconfiguration problem using the IEC 62443 standard, representing ICS as a network of zones linked by conduits. We present a system model incorporating security levels, device capacities, application dependencies, and communication constraints. Then, we formulate the task migration as an optimization problem solved via constraint programming. We detail several variations of the base reconfiguration program, including the activation of countermeasures or conduits, and the preemptive allocations of applications instances to host devices with memory size constraints. Our approach is evaluated through a combination of a physical training factory use case and generated problem instances with arbitrary sizes. This evaluation concerns the execution time of the reconfiguration process, as well as the resilience, measured in number of devices attacked before a critical application must be stopped.
The methodology and experimental validation were presented in 18. The instrumentation scripts are public 35.
9 Bilateral contracts and grants with industry
9.1 Bilateral grants with industry
Qarnot computing
Participants: Raphaël Bleuse, Kouds Halitim, Bogdan Robu, Eric Rutten.
We have a cooperation with Qarnot computing in the framework of the Inria challenge PULSE (website), with the support of Ademe, on the topic of pushing carbon-neutral services towards the edge. Particularly, we are involved in WP5 on the Control of emissions of intensive computation tasks, and WP6, which we are coordinating, on the efficient hybridation of heterogeneous computing tasks.
RTE
Participants: Stéphane Mocanu, Clément Mommessin, Eric Rutten, Jolahn Vaudey.
We have a cooperation with RTE (the French Energy Transportion company) : Guillaume Giraud, following our recent work in the H2020 CPS4EU project. It is continuing in the new project Tasting (Section 10.2.3) of the PEPR TASE.
10 Partnerships and cooperations
10.1 European initiatives
10.1.1 Seanergys
Participants: Raphaël Bleuse.
Since Septembre 2025, we are involved in the Searnegys project (website). Our involvment in this project is mainly in the work package 4, focusing on dynamic resource management. We develop simulation tooling for HPC platforms. This participation in the project strenghthen the existing collaboration with the Datamove team.
10.2 National initiatives
10.2.1 PEPR Cybersecurity, project SuperviZ
Participants: Omayma Alla, Gwenaël Delaval, Léa Kenmogne, Elian Loraux, Stéphane Mocanu, Eric Rutten, Jolahn Vaudey.
We participate in the PEPR Cybersecurity research project SuperviZ in three workpackages.
The SuperviZ project is granting four Phd's ins the areas of dectection and reaction to attacks (Figure 7).
The image presents the NIST Cybersecurity Framework, depicted as a circular diagram divided into five segments: Identify, Protect, Detect, Respond, and Recover, all governed centrally. Four PhD research topics related to cybersecurity are linked to different parts of the framework. PhD Elian Loraux (2028) focuses on OT Honeypot and Naval Group collaboration, linked to Respond. PhD Jolahn Vaudey (2025) on OT/IT reconfiguration and IEC 62443 DSL, linked to Respond. PhD Omayma Alla (2027) on Host IDS for PLC, linked to Detect. PhD Léa Kenmogne (2026) on Explainable AI for intrusion detection in ICS, linked to Detect.
Stéphane Mocanu is the leader of the Platforms workpackage of SuperviZ. An mid-term advancement report of the project is available 36.
10.2.2 PEPR Cloud, project Taranis
Participants: Raphaël Bleuse, Robin Chaussemy, Nathan Rabier, Eric Rutten.
In the framework of the PEPR Cloud, Ctrl-A is participating in the project Taranis (Model, Deploy, Orchestrate, and Optimize Cloud Applications and Infrastructure). We mainly work within WP 3 Orchestration of services and ressources.
The project is funding two PhDs working on autconomic managers. The focus is on the integration – as model and decision tools – of control and constraints, control and scheduling. The integration of temporal aspects in reconfiguration management will also be studied.
This work is done in cooperation with the Spirals team at Inria Lille and the Stack team in Nantes.
10.2.3 PEPR TASE, project Tasting
Participants: Stéphane Mocanu, Clément Mommessin, Eric Rutten.
In the framework of the PEPR TASE (Technologies Avancées des Systèmes Énergétiques), Ctrl-A is participating in the project Tasting (TrAnsformation of the energy SysTem for a better resilience and flexibility with enhanced digitalization), particularly in :
- WP1 : Infrastructure reliability and security with a post-doc position on : specification of the distributed architecture and reconfiguration strategy for the communication and control infrastructure
- WP2 : Distributed architectures of cyber-physical systems
with a
post-doc position on :
methods for attacks detection by events correlation
between network traffic observations and logs from control equipment. This postdoc is already ongoing and we start developping implementation tools for attacks on real-time protocols (SMV, GOOSE and PTP). The experimental validation is performed on G-ICS lab on a substation protection workbench. The electrical network is simulated by the G-ICS HIL system.
The image shows a mobile technical setup with three open cabinets. Each cabinet contains electronic equipment, an Intelligent Electronic Device (protection relay) and Stand-Alone Measurement Unit (SAMU) upper sections and the sets of electronic boards used to simulate the electrical network on the lower sections.
Figure 8: Substation workbench - WP3 : Ease deployment on hardware
with
post-doc positions on :
- model-based control (constraints solving) of self-adaptive deployment of distributed applications on the Cloud-Edge infrastructures
- reactive infrastructures for rapid protection in case of process perturbation
10.2.4 AMI CMA CyberSkill@Alpes
Participants: Stéphane Mocanu.
We participate in the AMI CMA Cyberskill@Alpes cybersecurity education and research project.
Stéphane Mocanu is involved into the AMI CMA Cyberskill together with a small team of technical staff not affiliated with CTRL-A. He is in charge of four deliverables:
- A demonstrator of IEC 62443 security approach deployment (zone/conduit definition and implementation). The demonstrator consists of five interconnected Fischertechnik Training Factory Industry 4.0 models twenty PLC and five industrial firewalls and a local network intended to show the security levels calculation, deployment of network flow cartography and network segmentation. The mechanical and electrical parts of the demonstrator are ready, we are starting the software development. A security supervision module will be added in the future.
- A low-cost workbench for cybersecurity awareness. Based on a simulated physical process, a low-cost PLC and HMI and license free software the workbench is intended to be used by high-school and secondary school teachers. We have designed the plans and acquired the hardware; the physical process simulation will be developed by computer science students in 2026 and the electrical and mechanical part of the workbench shall be achieved by the end of 2026
The image shows four FischerTechnik Factory Industry 4.0 simulators fully instrumented with PLS's industrial switches and firewals.
Figure 9: IEC 62443 demonstrator - A set of serious games for industrial systems security awareness and training. We intend to develop a board game with several complexity levels combining strategy, collaboration, resource management, hazard and cheating in a red team / blue team setup. A professional in game development was contacted and we hope to start the developments during 2026.
- An industrial systems security module including masterclasses and labs. Intended initially for the local cybersecurity master students the project has evolved to a training module that well be distributed by the PTCC (Programme de Transfert au Campus Cyber). A PTCC Formation grant was obtained in November 2025 and the developments are expected to start in January 2026. The labs will be developed on the Airbus CyberRange.
10.2.5 ANR RADYAL
Participants: Bogdan Robu, Eric Rutten.
Ctrl-A participates in the ANR project (in the ANR call : AI computing hardware architectures and accelerators in the context of Edge Computing) called Radyal Resource-Aware DYnamically Adaptable machine Learning, in cooperation with INSA Lyon / LIRIS (Stefan Duffner), the TARAN team, Inria / Irisa, Rennes (Marcello Traiola), and the MODUS team, UGA / GIPSA-lab, Grenoble (Bogdan Robu).
We will work on the analysis of self-adaptationreconfiguration spaces in the dimensions of Application (DNN algorithms), environment (applicative aspects e.g., lighting, obstruction in image analysis), and infrastructure and implementation configuration and deployment (involving hardware with reconfigurable precision and mapping). A post-doctoral fellow has started at GIPSA-lab on this project.
11 Dissemination
11.1 Promoting scientific activities
11.1.1 Scientific events: organisation
Member of the organizing committees
Stephane Mocanu is participating in the steering committee of RESSI (Rendez-Vous de la Recherche et de l'Enseignement de la Sécurité des Systèmes d'Information).
Eric Rutten is participating in the steering committee of FETCH (École d'hiver Francophone sur les Technologies de Conception des Systèmes Embarqués Hétérogènes) the Winter School on Heterogeneous Embedded Systems Design Technologies, for the 2023, 2024 and 2025 editions (Fetch web site).
11.1.2 Scientific events: selection
Chair of conference program committees
Eric Rutten was keynote session chair at IC2E25.
Member of the conference program committees
Raphaël Bleuse was member of the program committees of the PECS workshop at Euro-Par 2025 and the Symposium on Applied Computing (SAC) 2026.
Stéphane Mocanu was member of the program commitees of Netsoft workshop Secsoft 2025 and ESORICS workshop Anubis 2025.
11.1.3 Journal
Member of the editorial boards
Eric Rutten and Sophie Cerf were special issue co-editors for ACM TAAS (ACM Transactions on Autonomous and Adaptive Systems Special Issue on Control for Computing Systems).
Reviewer - reviewing activities
Raphaël Bleuse reviewed for the IEEE Transactions on Parallel and Distributed Systems (TPDS) journal.
Stephane Mocanu was reviewing for ACM computing surveys, Computers & Security, Electrical Engineering and Journal of Network and Systems Management.
11.1.4 Invited talks
Stéphane Mocanu was invited to a keynote on the cybersecurity of electrical grids at the SATES conference in 2025 in Nancy 19.
Clément Mommessin was invited for a seminar at the IRIT laboratory in Toulouse, in July 2025.
11.1.5 Scientific expertise
Stephane Mocanu is member of the Scientific and Technical Comitee of the SEE society and also of the Technical Commitee of the PTCC (Programme de Transfert au Campus Cyber).
11.1.6 Research administration
Raphaël Bleuse is member of the team organizing the LIG keynotes.
Gwenaël Delaval is elected member at the Academic Council (Conseil Académique) of University Grenoble Alpes (UGA) for the Confédération Générale du Travail trade union.
11.2 Teaching - Supervision - Juries - Educational and pedagogical outreach
11.2.1 Teaching
- Master : Omayma Alla , Computer Networks and Cybersecurity, 28h lab, M1, Grenoble-INP/ENSE3
- Licence: Raphaël Bleuse , Communication bas niveau, 4h tutorials, 6h practicals, L1, Univ. Grenoble Alpes/IUT2
- Licence: Raphaël Bleuse , Introduction aux services réseaux, 8h tutorials, 8h practicals, L1, Univ. Grenoble Alpes/IUT2
- Licence: Raphaël Bleuse , Installation de services réseau, 8h practicals, L1, Univ. Grenoble Alpes/IUT2
- Lifelong-learning: Raphaël Bleuse , Fondamentaux en informatique, 22h lectures, 8h practicals, Univ. Grenoble Alpes/IUT2
- Licence : Gwenaël Delaval , Bases du développement logiciel, modularité et tests, 15h lecture/tutorials, 15h lab, L2, Univ. Grenoble Alpes
- Master : Gwenaël Delaval , Sémantique des Langages de Programmation et Compilation, 30h tutorials, M1, Univ. Grenoble Alpes
- Master : Gwenaël Delaval , Compilation project, 4 weeks software project tutoring, M1, Univ. Grenoble Alpes
- Master : Lea Astrid Kenmogne Mekemte , Computer Networks and Cybersecurity, 28h lab, M1, Grenoble-INP/ENSE3
- Master : Lea Astrid Kenmogne Mekemte , Computer Networks and Automation, 10h class, 28h lab, M1, Grenoble-INP/ENSE3
- Master: Elian Loraux , sécurité de l'information: Red team Blue team DNS, 15h practicals, M2, Grenoble-INP/Ensimag
- Master : Stephane Mocanu , Computer Networks and Cybersecurity, 16h class, 34h lab, M1, Grenoble-INP/ENSE3
- Master : Stephane Mocanu , Industrial Computer Networks, 8h class, 8h lab, M2, Grenoble-INP/ENSE3
- Master : Stephane Mocanu , Reliability, 10h class, 8h lab, M2, Grenoble-INP/ENSE3
- Lifelong-learning : Stephane Mocanu , Industrial Control Sytels Cybersecurity, Inria Academy, One day State-of-the-Art Training.
- Licence: Clement Mommessin , Automatisation de la chaine de production, 6h lectures, 14h practicals, L3, Univ. Grenoble Alpes/IUT2
- Licence: Clement Mommessin , Virtualisation avancée 4h lectures, 12h practicals, L3, Univ. Grenoble Alpes/IUT2
11.2.2 Supervision
- PhD: Jolahn Vaudey (UGA), Self-reconfiguration of industrial systems applied to cyberresilience ; started October 2022, defended October 2025; co-advised by Stéphane Mocanu, Gwenaël Delaval, Eric Rutten.
- PhD in progress: Kouds Halitim (Inria), Efficicient task hybridization in heterogeneous computing: practical combinations of Control and Scheduling theories ; started November 2023 ; co-advised by Raphaël Bleuse, Éric Rutten, Bogdan Robu.
- PhD in progress: Lea Astrid Kenmogne Mekemte (UGA), Explainable AI for Network Intrusion Detection in Industrial Control Systems ; started November 2023 ; advised by Stéphane Mocanu.
- PhD in progress: Omayma Alla (UGA), Host Intrusion Detection Systems for PLC ; started January 2025 ; advised by Stéphane Mocanu.
- PhD in progress: Elian Loraux (UGA), Interactive honeypot fior industrial control systems ; started October 2025 ; co-advised by Stéphane Mocanu, Julien Francq (Naval Group) and Jean Leneutre (Telecom Paris).
- PhD in progress: Robin Chaussemy (Inria), Modelization of HPC Jobs and Resources to Minimize Energy Waste; started October 2025 ; co-advised by Raphaël Bleuse , Franck Corset (LJK) and Éric Rutten .
- PhD in progress: Nathan Rabier (Inria), Handling dynamic constaints and deadlines in distributed software reconfiguration ; started November 2025 ; co-advised by Sophie Cerf , Hélène Coullon (IMT Atlantique) and Eric Rutten .
11.2.3 Juries
Stephane Mocanu was reviewer and member of the PhD dissertation committees of Ahmed ELMARKEZ, U. Bretagne Sud and Nourhan HALAWI GHOSON, ENSAM, Nov. 2025 and also Enzo D'ANDREA, U. Lorraine, in Dec. 2025 and member of the CSI comitees of Arthur TRAN VAN, Telecom Paris and Florent DURECU (U. Clermont-Ferrand).
Eric Rutten was reviewer and member of the HdR jury of Hèlene Coulon (Univ. Nantes).
12 Scientific production
12.1 Major publications
- 1 articleA Domain-specific Language for The Control of Self-adaptive Component-based Architecture.Journal of Systems and SoftwareJanuary 2017HAL
- 2 articleModel-based design of correct controllers for dynamically reconfigurable architectures.ACM Transactions on Embedded Computing Systems (TECS)153February 2016HAL
- 3 articleDesigning Autonomic Management Systems by using Reactive Control Techniques.IEEE Transactions on Software Engineering427July 2016, 18HALback to text
- 4 inproceedingsSustaining Performance While Reducing Energy Consumption: A Control Theory Approach.Lecture Notes in Computer ScienceEURO-PAR 2021 - 27th International European Conference on Parallel and Distributed Computing12820Euro-ParLisbon, PortugalSpringerSeptember 2021, 334–349HALDOI
- 5 inproceedingsCombining neural networks and control: potentialities, patterns and perspectives.Proceedings of The 22nd World Congress of the International Federation of Automatic ControlIFAC 2023 - 22nd World Congress of the International Federation of Automatic ControlYokohama, JapanJuly 2023HAL
- 6 inproceedingsDiscrete Control of Response for Cybersecurity in Industrial Control.IFAC 2020 - IFAC World Congress 2020Proc. of the 21st IFAC World CongressBerlin, GermanyJuly 2020, 1-8HAL
- 7 articleIntegrating Discrete Controller Synthesis in a Reactive Programming Language Compiler.journal of Discrete Event Dynamic System, jDEDS, special issue on Modeling of Reactive Systems2342013, 385-418URL: http://dx.doi.org/10.1007/s10626-013-0163-5
- 8 inproceedingsReactive model-based control of reconfiguration in the Fractal component-based model.Proceedings of the 13th International Symposium on Component Based Software Engineering (CBSE), \em Prague, Czech Republic, 23-25 June\bf best paper award2010, 93--112URL: http://dx.doi.org/10.1007/978-3-642-13238-4_6
- 9 inproceedingsA Domain-specific Language for Autonomic Managers in FPGA Reconfigurable Architectures.ICAC 2018 - 15th IEEE International Conference on Autonomic ComputingTrento, ItalyIEEESeptember 2018, 1-10HALback to text
- 10 inproceedingsEfficient Mining of Temporal Safety Properties for Intrusion Detection in Industrial Control Systems.SAFEPROCESS 2018 - 10th IFAC Symposium on Fault Detection, Supervision and Safety for Technical ProcessesVarsovie, PolandAugust 2018, 1-8HAL
- 11 incollection What Can Control Theory Teach Us About Assurances in Self-Adaptive Software Systems? Software Engineering for Self-Adaptive Systems 3: Assurances 9640 LNCS Springer May 2017 HAL back to text
- 12 inproceedingsDevelopment Tools for Rule-Based Coordination Programming in LINC.19th International Conference on Coordination Languages and Models (COORDINATION)LNCS-10319Coordination Models and LanguagesPart 2: Languages and ToolsNeuchâtel, SwitzerlandSpringer International PublishingJune 2017, 78-96HALDOI
- 13 inproceedingsIAS: an IoT Architectural Self-adaptation Framework.ECSA 2020 - 14th European Conference on Software ArchitectureL’Aquila, ItalySeptember 2020, 1-16HAL
- 14 incollectionFeedback Control as MAPE-K loop in Autonomic Computing.Software Engineering for Self-Adaptive Systems III. Assurances.9640Lecture Notes in Computer ScienceSpringerJanuary 2018, 349-373HALDOI
- 15 articleAn Autonomic-Computing Approach on Mapping Threads to Multi-cores for Software Transactional Memory.Concurrency and Computation: Practice and Experience3018September 2018, e4506HALDOIback to text
12.2 Publications of the year
International journals
Invited conferences
International peer-reviewed conferences
Conferences without proceedings
Doctoral dissertations and habilitation theses
Other scientific publications
Software
12.3 Cited publications
- 36 techreportSuperviZ supervision et orchestration de la sécurité Rapport d'avancement à mi-projet.Télécom SudParis (Institut Mines-Télécom) ; InriaOctober 2025, 1-56HALback to text
- 37 bookFeedback Control of Computing Systems.Wiley-IEEE2004back to text
- 38 articleStandard specification-based intrusion detection for hierarchical industrial control systems.Information Sciences659February 2024, 120102HALDOIback to textback to textback to text
- 39 inproceedingsCorrupted GOOSE Detectors: Anomaly Detection in Power Utility Real-Time Ethernet Communications.GreHack 2015VerimagGrenoble, FranceNovember 2015HALback to text
- 40 articleThe Vision of Autonomic Computing.IEEE Computer361January 2003, 41--50back to text
- 41 articleCross-domain Alert Correlation methodology for Industrial Control Systems.Computers and Security118JulyApril 2022, 102723HALDOIback to textback to text
- 42 inproceedingsDetecting Process-Aware Attacks in Sequential Control Systems.NordSec 2016 - 21st Nordic Conference on Secure IT SystemsOulu, FinlandNovember 2016, p.20-36HALDOIback to textback to textback to textback to text
- 43 inproceedingsEfficient Mining of Temporal Safety Properties for Intrusion Detection in Industrial Control Systems.SAFEPROCESS 2018 - 10th IFAC Symposium on Fault Detection, Supervision and Safety for Technical ProcessesVarsovie, PolandAugust 2018, 1-8HALback to text
- 44 inproceedingsExperimental study of performance and vulnerabilities of IEC 61850 process bus communications on HSR networks.EuroS&PW 2020 - IEEE European Symposium on Security and Privacy WorkshopsGênes, ItalyIEEESeptember 2020, 584-593HALDOIback to text
- 45 articleReal-Time Performance and Security of IEC 61850 Process Bus Communications.Journal of Cyber Security and Mobility102April 2021, 1-42HALDOIback to text
- 46 articleOn the Supervisory Control of Discrete Event Systems.Proceedings of the IEEE771January 1989back to text
- 47 inproceedingsDesign Framework for Reliable Multiple Autonomic Loops in Smart Environments.2017 IEEE International Conference on Cloud and Autonomic Computing (ICCAC) Tucson, AZ, United StatesSeptember 2017HALback to text