Key people
In the news
- Recently, Leap was invited to give a keynote on AI for agricultural machines at the general assembly of Fedagrim. The part of the talk that resonated most in the room was pattern recognition. An LLM like ChatGPT recognises patterns in text and predicts what comes next. A model on an agricultural machine recognises patterns in drive pressure and ground speed and estimates the throughput. Both are AI. The one on the machine needs far fewer parameters, but it has to be robust enough to run all season. Pattern recognition is also what an
- Sometimes you need to feel the temperature of a machine with your own hands to understand what its data is telling you. This time, that meant a round trip to China. Four intense days on-site, working out why a machine's data behaves the way it does. And there's a real thrill in adding a new measurement and watching it suddenly line up with the effect you're chasing. Also spotted a namesake in the parking garage. We'll take it as a compliment 😉
- If every operator had the insight of an engineer, shutdowns would be a lot shorter. Nobody can wait around for one, of course. Depending on experience, people need to work through the possible causes until the machine runs again. Often the cause is already in the sensor data, and reading it is the part that takes the engineer. We built a root cause analysis model that reads the signal after a shutdown and gives the operator a likely cause to start from, with a confidence score per failure mode. Full walkthrough in the carousel. Happy
- Proud to share that we've successfully closed out the first stage of the Astromechatronics project with Astrofood, made possible by an ESA Space Solutions Belgium Spark Funding. Over the course of this project, we took important steps toward automated, intelligent systems for space-based microalgae cultivation, combining mechatronics, control theory, and data science to push this technology forward. Grateful for the collaboration and the trust from ESA Space Solutions Belgium, and looking forward to future opportunities to take this
- The real AI advantage in industry isn't always the algorithm: it's often the people who know the process better than anyone else. Our co-founder Tom Staessens shared his take with Agoria on the future of industrial AI in Belgium. Check it out below👇
- Most machine teams have someone like this: the person who just knows how to get the machine right. Everyone is glad they're there, and nobody enjoys the week they're on holiday. When a machine leans on one person's feel, that dependency turns up in more places than you might expect. We put three of them in the carousel below. Do you have a person like that on your floor? Which of these three worries you most?
- Our co-founder Tom was interviewed by Agoria, check out his take on industrial data in Belgium below 👇
- On most machines, the settings that decide the outcome live in one person's head. That works, right up until they're on holiday, or a new variant shows up, or the material starts behaving differently than it did last month. Meanwhile the machines get more configurable and the jobs more variable. More knobs, less time to turn them. So we put together a primer on how much of that a machine can work out for itself. What gets called setting automation mostly comes down to three moments: what past jobs tell you before a run, what
Something wrong or missing? Send an update. Fixed within 24 hours.