HyperCOG: Cyber-physical systems for industrial processes

Before developing its own innovative digital technology, the HyperCOG project began its work by identifying existing state-of-the-art digital solutions used in smart manufacturing as a basis on which to develop the HyperCOG solution. This work included analysing existing frameworks relevant for the three use cases involved in the project which are in the chemical, cement and steel industries. This early work also examined relevant ethical, legal, privacy and security requirements that are needed to keep the cyber-physical system (CPS) architecture cyber secure in the real world.

Based on the requirements obtained from this work, an innovative industrial cyber-physical system has now been developed, along with a hyperconnected CPS network made up of digital nodes and the communication architecture required to support industrial production. This architecture advances the concept of cognitive manufacturing, by combining cognitive computing techniques (such as artificial intelligence), the industrial internet of things, and advanced data analytics to optimise manufacturing processes.

“HyperCOG involves the latest advances in AI relevant to industry, such as modelling for twin factories and decision-support systems for human-machine interaction,” explains HyperCOG’s coordinator Beatriz Chicote of Lortek in Spain.

“As part of this development, relevant industrial data have been collected to investigate data analysis methods that can obtain the knowledge hidden in the data and make it usable for optimising processes that occur in the project’s three use cases – steelmaking at Sidenor in Spain, cement manufacturing at Çimsa in Turkey and chemical extraction at Solvay in France.

“As of now, a preliminary model along with optimisation algorithms have been developed ready for the optimisation of the production planning in the steel use case. Lessons learnt from this exercise will then be used in the implementation phases of the other two use cases,” she adds.

To provide further context to this HyperCOG solution, it is important to understand the HyperCOG node architecture, which ensures the CPS meets the technological requirements of Industry 4.0. It is these digital nodes that form the basis of the system’s hyperconnected network architecture, which is composed of nodes running in several devices that communicate with each other without hierarchical layers or the need for either gateways or message brokers.

Starting from the Reference Architecture Model for Industry 4.0 (RAMI 4.0), the system breaks down traditional hierarchical information systems, due to its hyperconnecting capabilities. The nodes can acquire outstanding streams of data in real-time, which together with high computing capabilities, provide sensing, knowledge and cognitive reasoning, which will ensure that the companies using HyperCOG systems are robust in the face of variant scenarios – the system adapts.

“To be able to acquire information in real time from data provided by a physical sensor connected to a PLC or other acquisition device or system, the nodes collect historic data, as well as record data, and executor nodes run models or algorithms to check the status of the system in real time,” says Chicote. “To do this, middleware has been used to abstract the communications in a distributed system, making it a cyber-physical system, composed of several connected devices.”

As well as this node-based architecture, HyperCOG provides compatibility with the Functional Mock-up Interface standard. This enables interoperability between different software languages to achieve interoperability between models or algorithms developed in different environments to build more complex models or algorithms being developed for different use cases.

Finally, a monitoring tool has been developed which monitors the acquired data, analyses the correct communication between the different nodes, monitors the status of these nodes, reports logs that indicate possible mistakes, and shows these in a colour scale to indicate the level of the error. This tool is also able to verify the value of data and to estimate how the system can be operated correctly.

HyperCOG is now looking to provide a comprehensive and quantitative sustainability assessment of this industrial cyber-physical system. The goal and scope definition of the LCA for the three case studies have already been documented and the analysis is currently under way at the first deployment of the solution at the Sidenor steel facility in Basauri, Spain.

“HyperCOG will show the potential of its technologies at each of the three use cases and will also evaluate their replicability and transferability to different industrial sectors and to companies of different sizes,” says Chicote. “Our objective is to increase and improve production performance while reducing environmental impact by reducing energy consumption and CO2 emissions.

“Not only will society benefit from this work through this environmental impact, but we will see further benefits through the development of strategies for training and re-skilling workers, and the vocational training of students, for the use of digital technologies in industry, both of which are important aspects of the project’s ongoing work,” she concludes.