The operation of processes involves a set of complex processes covering various operating modes (in production, stopped, starting, etc.) adapted to the criticality of the encountered operating modes (normal operation, incidental and accidental operation) according to different time constants (real-time operation, offline maintenance, . . .). Today, these processes are based on interactions between the different professions of operation (planners, operators, maintenance agents) and heterogeneous proprietary systems (digital control system, and generally limited to the normal production phase). The CISPI demonstrator illustrates the research work of the ISET department in the field of System Engineering and Functional Safety.
In the field of System Engineering, the platform serves as a support for the specification, design, and validation of complex systems integrating technical, organizational, and human components. The emphasis is on the development of models, methods, and tools for System Engineering. In this field, the platform has notably served as an experimental support for the work of Fabien BOUFFARON (thesis defended in 2016) on the co-specification and co-simulation of heterogeneous models (physical system, control system, command system, actuation and measurement systems, etc.) for the interactive operation of a critical industrial process.
In the field of Functional Safety, the platform serves as an experimental support for the design of safe operation integrating the dimensions of operation, monitoring, and maintenance. The first work initiated in the context of the thesis of Dragos DOBRE (2010) focused on the development of an Interactive Driving Aid System. The work of Thomas COCHARD (2017) coupled with the post-doc of Mohamed Bouazziz (2015) focused on the automatic generation and formal verification of driving procedures, as well as the development of a decision support system based on classification techniques of the admissible sequences obtained. The objective is to provide assistance to planners by proposing and classifying (according to performance, availability, health checks) action sequences allowing to reach an "objective" situation, while respecting the safety constraints related to the process. These works also contribute to the validation of architectures of critical industrial processes: one of the main challenges is to provide a guarantee that the designed architecture will be able to react safely to critical situations and events. Finally, the platform also serves as a demonstrator for research actions in the field of maintenance, prognosis, and PHM, notably through the tools developed by the company PREDICT
The process of the demonstrator is composed of a set of tanks, valves (23 manual and 7 pilot-operated), pumps (4), pipes, and measuring instruments (flow sensors, height sensors, . . .). It can be remotely controlled from a control room or locally, via a safety control/command architecture, based on Siemens hardware (Automate S7 315F 2PN/DP, remote I/O boxes ET200S and ET200SF, screen MP377).
The communication architecture is distributed over 2 networks. On the one hand, the "industrial" network interacts with the decentralized control equipment, and on the other hand the "enterprise" network where very diverse information circulates (teaching/research, accounting, internet). This strategy secures the exchanges between components of the platform against potential attacks (viruses, malware), and makes communications independent of the load of the enterprise network (downloads, videoconferences, . . .). An OPC (OLE for Process Control) server serves as a bridge between these 2 networks so that remote applications from the control room can supervise the process. The "industrial" network consists of heterogeneous solutions (RFID, Wifi IEEE 802.15.4, Profisafe, Wifi 802.11.g and wired Ethernet) to communicate the different "intelligent" components. In order to ensure the interoperability of the components, several gateways have been developed and/or used. The Crossbow Stargate Net Bridge gateway ensures the return of information from the MICAz wireless sensor network to the supervision. The implementation of new Information and Communication Technologies (OPC server, industrial VLAN, CRAN VLAN, web servers in the APIs, . . .) opens up the automation of the platform to the enterprise network (control room, ERP, etc.) and to the Internet (e-supervision).
Software tools :
— System engineering : IBM Rational Rhapsody ;
— Control/command engineering : Siemens Step 7, WinCC, iMap ;
— Wireless sensor networks : Crossbow MoteConfig, XSniffer ;
— Simulation : Dassault Systèmes Dymola ;
— PHM : Predict Suite Casip/Kasem.
