As a machine builder, we regularly get the question, “Do you also do anything with AI?” The honest answer: yes. But not in the way you might expect.
Beyond the hype
AI is everywhere. In headlines, at conferences, in every startup's pitch. We are not an AI company. We design and build machines. But the question is not whether you work with AI as a company. The question is how thoughtfully you do it.
At SMO, we deploy AI on two levels: in our own work processes and in the machines we build. Not because it's trendy, but because it helps us do better work.
In our machines: eyes that don't get tired
Vision systems are not new in machine building. Cameras that scan products, detect deviations, sort by colour or shape. We have been doing that for years. What has changed: the intelligence behind those cameras.
In a recent packaging control machine, we use a combination of RFID and camera recognition to automatically identify and sort incoming packaging. You used to programme something like this with hard rules: “if pixel X has this colour, then action Y”. Now we train systems on examples. They learn to recognise patterns that we cannot always describe ourselves.
The result: machines that handle variation better. A slight discolouration, a slightly different positioning, a new type of packaging: an AI-controlled vision system adapts where a rule-based system gets stuck.
We integrate this kind of technology where it makes sense. Not as a feature to sell, but as a means to build machines that do what they are supposed to do. Reliable, even when conditions vary.
In our work processes: from experiment to structure
At the other end of the spectrum, we use AI in our own operations. Take this example: the complete repositioning of our website, from ‘machine builder’ to ‘design-build partner’, was done in close collaboration with AI.
That doesn't mean a chatbot writes our texts. It means we use AI as a sparring partner, researcher and first writer. The nuance, the knowledge of our projects, the tone of voice: those come from humans. But the preliminary work, the structure, the first versions? That's where AI helps enormously.
Why do we tell you? Because we believe that transparency about this should be normal. Output depends on the quality of the input and the critical eye of the people working on it. AI does not replace craftsmanship. It accelerates it.
What we are investigating now
We are currently exploring whether we can build specific tools with AI support. Think of a system that helps prepare quotes for one of our specialised services. Not by guessing prices, but by cleverly combining historical data, component choices and project parameters.
That sounds more ambitious than it is. Many companies struggle with the same question: how do you deploy AI in a way that really adds value, without getting bogged down in technology for technology's sake? We experiment, evaluate, and only build on what works.
No magic. Added value, though.
Want to discuss a specific challenge? Get in touch.