Programming welding by hand: A few years ago, when I was still at Siemens, I remember Wandelbots visiting us and showcasing their Trace Pen: you could teach a robot by simply waving a pen. Do you guys remember, Marco and Christian??? They’ve since pivoted toward a robotics OS, but it’s interesting to see others still pursuing that same vision of intuitive, hand-guided programming. Glance Vision Technologies Srl is applying this idea to welding with DARDO. ✅ Faster workflows, guide the robot directly instead of coding ✅ Precise motion capture, ideal for complex welding tasks ✅ Real-time feedback, fine-tune paths in seconds Website: https://lnkd.in/dDH3afwy Can hand-guided programming help skilled coders achieve more in less time while keeping precision high? Seen at Piotr Stanecki 🌐 —- Weekly robotics and AI insights. Subscribe free: scalingdeep.tech
Ilir Aliu thanks for sharing and mention me 😉 This Teach- Pen approach is nice, but in practice, when it comes to selecting welding parameters or communicating certain signals with the surrounding peripherals, such as a part turner, a rotary table, or a linear axis where the robot is mounted it fail. Such add-ons or teach-in aids reach their limits and unfortunately offer far too few advantages in practice. There are existing more promising solutions, such as AI-vision-based robotic solutions, where no programming or robot teaching is required at all!
This is a cool gimic, but will never be used in real life scenarios for welding. Using 3D vision to inspect a workpiece and then having the operator select which vertices to weld or map the weld from a CAD file, that's going to be the future of robot welding and... It is already here.
Fascinating approach, Ilir! Hand-guided programming really seems to bridge the gap between skill and efficiency. Do you think this will become standard in more industries soon? Your tone fits keen to link
Fast learning Robots without problems with parameters. Impressive
Excellent tech. I have peculiar interest in this fild. Can you tell the name of the product or research paper or technology name so I can learn this?
Many of the current welding robot/cobot systems now have seam/joint tracking capabilities. The pen is now out of a job.
Excel macro feature of robotics world can let robots learn by mimicking the moves faster , kids are very good at this, a valuable learning technique
Thank you for sharing and #tagging me Ilir Aliu! Glance Vision Technologies Srl did a really good job preparing solution that can really make a difference if it’s coming to welding! (and more)
We've demonstrated something similar in surgical robotics with a cooperative control system. Activities like inserting and extracting an endoscope during skull base surgery isn't just a straight in and out. Having the ability to record those dodges and undulations can greatly speed up the process of extracting the endoscope to clean off the tip.
Results-Focused Investor | Strategic Advisor. I turn big ideas into unstoppable ventures that scale fast. I talk about AI, Robotics and Growth
2dThe Wandelbots pivot is interesting from a capital allocation perspective. Hand-guided programming absolutely solved a real problem - teaching robots shouldn't require a CS degree. But moving to the OS layer makes sense when you're looking at where recurring value gets captured across the entire robotics stack versus just the teaching interface. DARDO sticking with the original vision isn't wrong though. For a smaller player, owning a specific application vertical (welding) might actually be smarter than trying to compete at the infrastructure layer where you need massive scale. Skilled coders can definitely move faster with hand-guided tools, but I'm more curious about the business model durability. Wandelbots made their bet on platform economics. DARDO is betting on application depth. Both could be right for different reasons.