Robot Learning Software Improves Robotic Spot Welding Cycle Time

FANUC’s patented Learning software improves robot performance by learning the characteristics of programmed paths and then speeding them up, all while maintaining the programmed paths and smoothing motion.

In this video demonstration of the FANUC Learning Robot, a FANUC R-1000iA/80F Learning Robot on the right competes with a standard R-1000iA/80F robot without the learning capability on the left. At the start signal, both robots execute the same spot welding program on two identical automotive panels. The robots “race” to complete the process in the fastest time. 


The demonstration also highlights FANUC’s new intelligent interference check, which automatically determines shared workspaces to avoid collisions. Intelligent interference check is extremely valuable and cost-effective in high-density cells where multiple robots work together.

The robots press their “Finish” buttons when their cycle is complete, and their cycle times are recorded. The Learning Robot shows a significant improvement in cycle time compared to the standard robot; Motion cycle time can be reduced by up to 15% with the added benefit of smoother overall motion. In addition to the Learning function, this system is equipped with FANUC’s new servo tip dresser control used by the robots at the start of a new racing cycle. This feature allows traceability of dress records and trends and eliminates the need for costly hardware. 

FANUC America’s Learning Robot is helping manufacturers increase production. Whether it’s eliminating a system bottleneck, or increasing overall line rate, the Learning Robot can help without large investments in capital equipment or the need for expert programmers.

Source FANUC America