When I moved from DSM to Covestro on 1 April 2021, I did not only move to another company. I also brought with me a small piece of the DSM world.
For the past five years, I have managed a number of MES implementations (computer systems for production management) that had originally been developed within DSM. The systems were running in manufacturing plants that moved from DSM to Covestro, and I moved with them.
These systems were never really part of Covestro’s standard MES landscape. They were, in a sense, legacy systems from the moment I arrived. There was therefore a fairly clear expectation that they would eventually be replaced by a Covestro solution.
“Eventually”, however, turned out to be a rather elastic concept.
In the meantime, the systems continued to do useful work. They provided historian data, but their role went considerably further. We collected batch-performance data from daily operations and used those data for Continuous Improvement. Within Covestro’s Analytical Platform, the data could be summarized and transformed into analyses that provided pointers to further process improvements.
There was also a less glamorous but very practical benefit. By automating recurring reporting and analyses, the systems took work away from local engineers and management. Some of the activities were simply tedious: producing regular reports, preparing statistics, or doing things such as annual target setting. Once the relevant data were available in a structured form, much of that work could be automated.
So although the systems were “non-standard”, they were not exactly redundant.
Running my own shop
What made my role interesting was that I was not simply the application administrator of an existing corporate system. Over time, I effectively had my own little “shop” within Covestro.
I was responsible for the whole environment: costs, IT interfaces, Managed Operations, migrations, extensions to the functionality, and all the practical things needed to keep the systems running and developing.
There was no large team behind me doing these things. In practice, I had to find the people I needed, arrange the interfaces, manage the suppliers and operations, decide what needed to be changed and make sure that the result worked.
Looking back, “intrapreneur” is probably a better description of what I was doing than “system manager”. I operated inside a large company, but with considerable independence. I had a small domain that I was responsible for and, within that domain, I had to make things happen.
Of course, that independence also meant that I sometimes felt rather detached from the mainstream Covestro organisation.
The museum
There was a particular place where this became almost a running joke.
In Geleen, there were still quite a few people around who had worked at DSM before the transition. Some were now at Covestro, while others had moved with parts of the old DSM organisation into successor companies.
During breaks and lunches, we would sometimes compare experiences. We would wonder about things that had changed after the transition, discuss things that we found difficult to understand, and, perhaps most importantly, laugh about them.
That laughter was useful. When something frustrates you, being able to tell the story to someone who immediately understands why it is frustrating makes quite a difference. And when you can laugh about it together, the frustration becomes easier to handle.
At some point I started referring to my little MES environment as my “museum”.
It was an appropriate name. The systems had their roots in an earlier DSM era, and I was still running them years after the organisation around them had changed.
One former colleague, now working at Envalior, took the joke a step further. Whenever we met — sometimes with other people around — he would address me as “Museum Director”.
I found that funny.
It also gave me an opportunity to explain what my “museum” actually contained. And the more I explained it, the more obvious it became that it wasn’t really a museum at all. It was still doing useful work.
There was something interesting about that wider Geleen community. It was not only a community of people from the same company. It included people who had ended up in quite different organisations but who had shared the same DSM background. Envalior is an interesting example: its engineering-materials business combines former DSM engineering plastics with the former LANXESS high-performance materials business. They too have been going through a transition in which an old organisational culture meets a new corporate environment. Maybe even similar, because LANXESS is also a Bayer successor, like Covestro.
We therefore had something in common that was difficult to describe in organisational charts. We recognised similar changes in the way work was organised, similar differences in expectations, and sometimes similar frustrations — even though we were no longer colleagues.
That community turned out to be valuable.
When a legacy system becomes useful
Meanwhile, the replacement of my “museum” was progressing — or, more accurately, being discussed for quite a long time before it actually progressed.
For years, the systems were regarded as something that should eventually disappear because they were not the standard. Eventually Covestro did develop a replacement with similar functionality and started rolling it out, including to sites that were still using “my” MES solution.
The plan is now to retire the legacy systems by the end of 2026.
There is a certain irony in the timing.
After years of having systems that were considered non-standard, the standard replacement has finally arrived. And it arrives just in time: in about six months, my retirement will lead to the wind-down of my activities for Covestro anyway.
The “museum” is therefore finally closing.
But the work continues
I am not completely leaving the story behind.
Within the replacement project, I still have a role in developing the new data platform. In essence, this is a 2.0 version of what I have been running.
The new platform goes a step further. It provides ready-to-use datasets for batch and batch-step analyses, including things such as duration histograms and statistical analyses. These data can then be used to derive manufacturing and scheduling targets.
That is actually the part I find most satisfying.
The systems themselves are technology. Technology has a lifecycle, and eventually old systems should be replaced. There is nothing wrong with that.
What matters more is what you do with the data — and the way you think about using them.
So, after all those years, my systems will probably not survive me.
But there is a fair chance that the way I worked with the data will.
And perhaps that is a better legacy for a Museum Director anyway.