Selected Project Results

Based on our analysis, we can provide insight information about past research and point towards important areas for future activities.

Our Database Access

Impact Dimensions

How do we assess impact? We use the statements provided in the projects final reports and an independent vote on the probability how the associated impact dimension was in fact targeted. We do not double penalise or double count here. Most projects did astonishingly well by achieving impact at plants. Some dimensions of impact were though underrepresented, as shown in the analysis result below.

Dimensional analysis yields a clear disadvantage of worker safety and worker performance related impact. Compared to quality improvement, which was by far impacted by the highest number of projects, enabling technologies, power consumption and emission reduction are considered to be underrepresented as well.

Assorted, Semantically Induced 
RADAR Charts of Projects and Methods

Impact analysis as function of time

 With our evaluation, we can not only map the impact with respect to methodology, but also against time. Here, we use the finalisation date of a project to show when a certain impact dimension was actually realised. 

Use the slideshow below to swipe through the different impact dimensions for getting the time dependency of the associated research projects.

We can see that the EU RFCS programme gradually financed research in various environmental targets like waste reduction, emission reduction and power consumption. From the numbers, more "green" projects have been funded since 2014, with peaks that can be observed in 2018. 

Worker safety and worker performance related research is underrepresented, but projects that did in fact consider those aspects are among the most recent ones.

Quality improvement and customer satisfaction were more in the focus of research projects during the early 2005 - 2010 and were slightly reduced in recent years

Introduction of enabling techniques into the context of automation and control for downstream steel processing was only marginally present and slightly increased since 2016. This is due to trending digital topics like Industry 4.0, Big Data and Machine Learning. 

Issue and barrier analysis

Projects never run perfect, if they do this is often a lucky exception. Research projects often require a "living plan" and frequent correction of course. They encounter issues that were not foreseen, force majeure events that lead to project delays.

Nearly every project suffered more or less minor delays. Perfect conductance is an exception, although this was indeed seen with at 5 evaluated projects. There was yet a relationship between the considered aggregate, the picked method and the encountered barriers and issues. The Sankey plot above represents one example of our analysis for the problem area rolling mill (note, the wider range of application includes hot-rolling, plate rolling, cold-rolling here).

Whenever technical components were relevant to the project, especially when shipment of parts was involvement, project delays were mostly due to delayed shipment, availability or technical functioning of this part. Example: Availability of IMPOC system took very long. Once available, the system was technically not connected to the plants IT infrastructure which took again some time.

Solution Space

Projects followed specific strategies and introduced solutions to tackle their technical problems. The solution space is composed of those advanced automation and control strategies that were found to be among the most influential to the former research projects and methods that experts would expect to find in similar settings. Therefore, ControlInSteel was able to determine not only the usage of methods but also to find methods that were neglected so far and which could be interesting starting points for future projects. 

Considering the choice of solutions, advanced automation and control projects focused on a dedicated set of techniques. We took the encountered methods and added those techniques that are similar or comparable. Then, each project report was checked for the applied methods and the projects were allocated to solution terms.

It is not surprising to frequently encounter internal model control and model-predictive control in our project portfolio - many projects were devoted to introduce these techniques to the steel processing domain and we cover a time frame, where these approaches were on its application high. This also explains the low number of classical control solutions like PID controllers. 

Nevertheless, projects rarely adopted mathematically more challenging concepts like Laplacian or Laurent transforms. The Koopman space theory is completely missing in the project coverage as a solution. Use of Bayesian statistics or challenging stochastic approaches is completely missing.

Project Vectorisation in Taxonomy Spaces

Each project can be mapped into the taxonomical space. In the following example, we focus on a project that optimised the refinement and increased yield. We also restricted the solution space for better visualisation.

How does our taxonomy work?

All processes are defined by their channel of interaction. For instance, roughing forms the product. It uses mechanical forming interaction which also exhibits a thermodynamical footprint on the product.

Mindmap visualisation of the problem space

The diagram summaries the results of the problem space taxonomy and distinguishes mainly chemical (red) and physical (green) interaction channels.

Problem Space Project Distribution

What problems were treated by research projects in the past? We wanted to get an quantitative view on the relative distribution of these topics. Below, you can see a block chart that represents the number of projects associated with a specific topic.

Distribution of projects onto different parts of the problem space. We see that long products have been for more rarely treated by research project than flat products in the past. Although we see a rising number of long product projects in recent years. 

An overwhelming part of the projects dealt with optimising the finishing mill.

Those aggregates of primary interest for the European Green Deal have not yet been frequently worked on: Galvanisation, Pickling Link, Annealing Line, Coating. Therefore a clear recommendation would be to drive research into that direction.