TLDR
- ABB Robotics is integrating NVIDIA’s Omniverse libraries into its RobotStudio software to close the “sim-to-real” gap in industrial robot training.
- The new product, RobotStudio HyperReality, promises up to 99% accuracy when moving from virtual training to real-world deployment.
- ABB says the technology can cut setup and commissioning times by up to 80%, reduce costs by 40%, and speed up time-to-market by 50%.
- Foxconn is already piloting the technology in consumer electronics assembly, with a full rollout to 60,000 RobotStudio customers planned for H2 2026.
- California-based robotic workforce company WORKR will demo the solution at NVIDIA GTC 2026, running March 16–19 in San Jose.
ABB Robotics announced on Monday it is partnering with NVIDIA to tackle one of the oldest problems in industrial automation — getting robots to behave in the real world the way they do in simulation.
The Swiss robotics company will integrate NVIDIA’s Omniverse libraries into its RobotStudio programming and simulation platform. The result is a new product called RobotStudio HyperReality, due for full release in the second half of 2026.
The core issue the partnership targets is known as the “sim-to-real” gap. For decades, simulations have struggled to replicate real factory conditions like lighting, shadows, textures, and physical variation. That mismatch has forced manufacturers to spend extra time and money bridging the two worlds.
ABB says its solution closes that gap with up to 99% accuracy. It is the only robot maker with a virtual controller running the same firmware as its physical hardware, which helps align simulation results with real-world performance.
The company’s Absolute Accuracy technology also reduces robot positioning errors from 8–15mm down to around 0.5mm, making it suitable for high-precision work like electronics assembly.
What the Technology Does
Manufacturers using RobotStudio HyperReality can design, test, and optimize production lines virtually before moving anything to the factory floor. ABB claims this can cut setup and commissioning times by up to 80%.
Cost reductions of up to 40% are also projected, largely by removing the need for physical prototypes during development. Time-to-market for complex products can be cut by 50%, according to ABB’s own analysis.
The system uses synthetic data to train robots on multiple tasks and production scenarios. Once trained virtually, robots can be deployed to the production line with the high accuracy the company promises.
ABB is also exploring the integration of NVIDIA’s Jetson edge computing platform into its Omnicore controller, which would enable real-time AI inference directly on the robot.
Early Adopters Already Testing
Foxconn, the world’s largest electronics contract manufacturer, is the first company piloting the joint solution. It is using RobotStudio HyperReality to train assembly robots for consumer electronics production — a task that involves delicate pick-and-place operations across multiple device variants.
Foxconn’s Chief Digital Officer, Dr. Zhe Shi, said the level of accuracy and fidelity the tool delivers “just wasn’t possible in simulation and digital twins” until now.
WORKR, a California-based robotic workforce company, is also using the platform. At NVIDIA GTC 2026 in San Jose (March 16–19), WORKR will demo AI-powered robotic systems built on ABB technology that can be operated without any programming knowledge.
WORKR CEO Ken Macken said the collaboration is focused on making industrial AI “deployable today,” particularly for small and medium manufacturers facing labor shortages.
ABB said RobotStudio HyperReality will be available to all 60,000 of its existing RobotStudio customers when it launches in the second half of 2026.
ABBN stock was down 4.22% and NVDA was down 3.01% at the time of reporting.





