ABB’s Robotics business unit has entered into a collaboration with NVIDIA, aiming to narrow the gap between how industrial robots perform in virtual simulation and their actual operation in factories.
Under the partnership, Switzerland-based ABB will adopt NVIDIA Omniverse simulation databases to make its robot training environments more realistic by incorporating details such as lighting, shadows and textures.
Marc Segura, President of ABB Robotics, stated that robots typically have limited perception of their surrounding environment, which affects their operational accuracy, repeatability and speed.
Segura cited an example: if robots work near stamping equipment in a factory, the strong vibrations generated during stamping can degrade robot performance.He told Reuters that in the past, robots had to learn gradually or be programmed to adapt to vibration effects.With this new technology, however, robots can be trained in a virtual environment and “have those capabilities from day one,” which “will save companies significant time and costs.”
The partnership reflects a growing industry trend where companies increasingly use digital simulation for production planning and robot deployment, identifying issues in advance before equipment is formally commissioned.
ABB said the system will be delivered through its robot control software, reducing the need for physical prototypes of products and production lines, thereby lowering costs and accelerating time-to-market.
The technology is scheduled to launch in the second half of 2026 and is expected to serve customers in the automotive, consumer electronics and other industries.
Foxconn, the electronics manufacturing services provider, has already piloted the technology for installing side buttons on consumer electronic devices.ABB noted that such tasks have traditionally been challenging due to shadows interfering with robot vision.
Deepu Talla, Vice President of Robotics and Edge AI at NVIDIA, said: “The industry requires simulation with high physical accuracy to break down the barriers between AI-driven robot training in virtual worlds and large-scale deployment in the real world.”
(Reprinted from https://news.eccn.com/）