Angelo Hulshout is an experienced independent software craftsman and a member of the Brainport High Tech Software Cluster.

31 March

Angelo Hulshout has the ambition to bring the benefits of production agility to the market and set up a new business around that. His journey is taking a small detour.

In 1998, a not so obscure metal band from ’s-Hertogenbosch released an album called “How to measure a planet.” The band was called The Gathering, and this album was a break away from their previous albums, which were filled with rather bombastic heavy metal. Daring, but they did it – and they were successful.

With a startup, it’s no different. At some point, you may have to deviate from your original plan to become successful. I remember the story of an ex-colleague at Philips who started his own company in compiler optimization. He ended up doing completely different things, in the area of static code analysis. The core technology he used stayed the same, but he wound up in a completely different market.

You don’t make a switch like this from one day to the next. It requires a bit, or a lot, of trial and error to get there. This is what we’re facing with Shinchoku right now, shortly after we started and even before the first version of our website is live.

Feet on the ground

My original idea is still alive: help small and medium-sized manufacturers optimize their production processes in a data-driven manner. Or in order words, make them part of Smart Industry, Industry 4.0. However, we now know that for this market, the world looks a little bit different than all the fancy Industry 4.0 videos show. There are very few robots, there’s still a lot of manual labor and although we live in the age of digitization, a lot of factories have very little production management software running.

That isn’t necessarily a bad thing, but it does require us to define which steps we want to take to realize our original goal. We’re not by default going in to install an internet gateway, hook it up to the PLCs of all production machines and then feed the data acquired to a dashboard. That was a bit of a short-circuited version of our story anyway – you can’t just collect random data and be successful, so we’d have to analyze with the customer what data to collect before taking off.

That analysis has put us with our feet firmly on the ground now: in most factories we’ve seen so far, there’s very little data to collect initially. We have to work with the customer, with the operators in the factory, to identify which parts of the process are candidates for improvement. This has to be done based on anecdotal information and experience because there’s only limited automation and reporting support. Once identified, we need to establish what can be done to (semi)automatically collect and arrange data before we can start working on our actual goal.

Small steps

At one factory, for example, about 60-70 percent of the production process turned out to be based on manual labor. Not only are the machines operated by humans, but the metal parts produced by those machines are also polished by hand. The problem in this factory is that nobody at any point during the production has an overview of how many parts for a certain order are being processed and where they are.

In a place like this, going from manual labor to full automation in one big bang is nothing more than a dream. The reality is that we have to take small steps. The first step, in this case, could very well be installing wireless foot pedals with certain workstations, which the operators can simply press every time they finish working on a single part. Combined with tracking the trays transporting the metal parts through the factory (using RFID or similar), that would help us get a better view of how many parts are produced and in which production stage they are. Obviously, this is lightyears away from just installing a few network cables and a gateway and collecting data that’s already there.

We discussed our experiences with both customers and automation specialists from our side of the industry. They all confirm that some important steps need to be taken first if we want Industry 4.0 to come within reach of the smaller production companies as well. Not everybody is as big as L’Oreal or Volkswagen, where millions can be spent each year to improve automation and human labor in production has been largely eliminated long ago.

Process analysis

In keeping with the song from The Gathering, we need to find out how to measure a planet before we can actually go ahead and measure it. That’s what we’re doing now. Realizing our Industry 4.0 vision requires some big steps, the first of which need to be taken in our market of small and medium-sized manufacturing companies.

So, almost from the start, Shinchoku will not only be a data analysis company, but it will also have a lot of work to do in process analysis and automation – almost literally to be able to generate its own data.

Edited by Nieke Roos