March 30, 2023
We spoke to Mikel Ayuso of LORTEK about the progress being made at the SIDENOR use case, where HyperCOG has been involved in steelmaking planning optimisation and ladle tracking.
Can we start by you introducing the SIDENOR use case in the HyperCOG project?
The SIDENOR use case focuses mainly on two different objectives: Steelmaking planning optimization and Ladle Tracking.
The aim of this use case is the optimization of the scheduled production process through the digitalization of the process and to solve on-line and offline production planning problems caused by unexpected events during steelmaking operations, which causes delays on the scheduled production. The current steelmaking planning for the week is done manually with the knowledge of SIDENOR’s process experts, to help these experts, the final objective here is to have a decision support system that will allow them to make better and faster decisions.
In this use case, the CPS (Cyber-Physical System) gets the data from the SIDENOR ERP (Enterprise resource planning) software and production environments, processes and analyses it. The output generated in the algorithms is shown to the workers through a new HMI system, this is both a real time representation of where the Ladles are, and a new plant sequence optimised to reduce scrap cost as much as possible. We essentially built a complete decision support system, with all this information the workers can decide if the optimizations calculated by the CPS are worth applying or not.
Digitising the process planning of the steelmaking shop was one of the main purposes of SIDENOR use case. How has this developed so far?
All the partners involved in the project are doing an amazing job, and this task is no exception.
A first complete version including the process optimization algotihms, the configuration and monitoring tools, the new HMI and the node based CPS itself has already been developed and is being tested with historical data provided by SIDENOR. The results that the CPS is showing using this historical data are very promising and make us believe that the workers from SIDENOR will find this tool interesting and useful, both to create new plannings and to optimize those that have been done manually.
The prototype to be installed in SIDENOR is aiming to reduce the number of production planning problems arising from failures that occur during the steelmaking process. How much progress has been made so far in the run up to the first implementation?
Due to the Covid-19 situation during the project, we unfortunately haven’t been able to deploy a first version on production yet, although our system is constantly being tested both in our own laboratory and various other partners test setups, all simulating the production environment with historical data provided by SIDENOR. All the required nodes are already working in all the test scenarios, and we have an iterative approach to our testing, continuously improving and optimizing the architecture and algorithms.
For April 2023 we expect to install the HyperCOG prototype in SIDENOR, which will work on their production environment for the last 3 months of the project.
How does your work tie in with the work being done in the project?
I am a researcher at Lortek and in the HyperCOG project, I am responsible for the implementation of the steel use case. I have been involved in this project since late 2019 as part of the team that has developed the CPS and all the nodes and communications for the three use cases. I have also worked in the integration of the AI algorithms into the CPS, as well as the connection with external actuators like the data protocols for capturing data and connection with the HMI frontend systems developed by other partners.
What have you enjoyed the most about being involved in HyperCOG?
The HyperCOG project works with three very different use cases and people with very different backgrounds and knowledge. This brings the opportunity to build something truly unique, a completely new development flexible enough to work in such different production processes, while also personally giving me the opportunity to learn a lot in the process, not only about my specific area of knowledge, but also about completely different topics.
Are you confident that HyperCOG will achieve its ambitious aims?
The collaboration between all the partners has been vital to advance in the different tasks and achieve the final objective. I firmly believe the team working on this project has done and is still doing an outstanding job, we have built a robust yet flexible solution that can work in very different use cases, and I can’t wait for the people working on production to finally get their hands on our developments, I believe their work will greatly benefit from the work done in HyperCOG.