Bizit Systems - Universal Robots

Inspection results shouldn’t depend on the time of day.

That drift usually isn’t the measurement but fatigue. It's loading parts into a CMM. Running the same gauge checks. Holding a scanner steady for hours.

This is where cobots often fit in 🔎 Supporting quality teams in repetitive part handling while inspectors focus on judgement.

More on where this works 🔗 http://urrobots.com/j4q

📹 Application by 3D Infotech

When your business model depends on customization, uptime matters.
THEMAGIC5 has relied on UR3e collaborative robots for more than seven years of non‑stop operation, producing thousands of unique goggle cuts with consistent precision. The company has scaled from Kickstarter to selling nearly half a million goggles worldwide 🦾

"When we produce a thousand goggles, it’s a thousand different face scans, and a thousand different cuts, provided by the UR3 robot," says CEO Rasmus Barfred

👉 Read the full case story http://urrobots.com/j3h

Training robots for insertion, assembly, polishing or handovers isn't just about where to move but what happens at first contact.

VP of AI Robotics Products Anders Billesø Beck talks about why torque-level access is critical for imitation learning that makes it out of the lab to survive on the factory floor.

Direct Torque Control allows learning systems to work directly with:

🔹 Joint torques
🔹 Calibrated robot dynamics
🔹 Real physical response at the end effector

This capability underpins our , where leader-follower demos capture synchronized motion, vision and force (on the same robots deployed in production).

👉 Read the full article http://urrobots.com/j3H

📹 Partner Generalist's packaging demo shows how advances in data collection translate to real-world performance

Most imitation learning works until the robot makes contact. Why? Because models are mainly trained on what the robot sees 👀 and where it moves.

Real industrial tasks aren't determined by trajectories but rather interaction. Insertion, assembly, finishing and handovers depend on:

- How force ramps when parts meet resistance
- How compliance is handled under misalignment
- How the system reacts when reality doesn't match the demo

If force and torque aren't captured during imitation learning, the robot never learns the operators physical intent only the path. This is why torque and force feedback matter 🦾

In his latest piece, VP of AI Robotics Products Anders Billesø Beck lays out how won't scale on perception alone.

👉 Read the full article http://urrobots.com/j3t

Inside a Brooklyn loft, UR3e collaborative robots perform thousands of different trimming operations on swimmer’s googles every week.

Integrated directly with THEMAGIC5's proprietary software, each robot executes a unique hot‑wire cut based on an individual face scan. The flexibility of the platform makes non‑repetitive, data‑driven manufacturing possible at scale.

A clear example of how robots enable customization without compromising speed or reliability.

👉 Read the full case story http://urrobots.com/j3f