The problem was that the robot spoke a language she couldn’t yet write efficiently. She could use the bulky teach pendant—a handheld controller tethered to the robot—but that would take weeks of stopping and starting, jogging axis by axis. Her deadline was in a month. She needed a faster way to design, test, and debug the robot’s movements without ever touching the physical machine.

In the early spring of 2019, a manufacturing engineer named Elena sat in her cubicle, staring at a dusty, 10-year-old industrial robot arm in the corner of the prototyping lab. The company had just acquired it from a decommissioned assembly line. Her task: reprogram it to handle delicate composite materials for a new aerospace contract.

Over the next week, she built a virtual cell: imported a CAD model of the composite part, positioned the robot model (an IRB 4600), and added a gripper from ABB’s library. Using , the robot’s native programming language, she wrote path routines, watched for collisions, and fine-tuned cycle times—all without wearing earplugs or risking a single real-world crash.