Key Highlights
- ABB Robotics incorporates NVIDIA’s Omniverse libraries into RobotStudio software to address industrial robot training accuracy challenges.
- RobotStudio HyperReality delivers up to 99% accuracy when transitioning from virtual environments to physical factory floors.
- The platform reduces setup and commissioning times by up to 80%, lowers costs by 40%, and accelerates time-to-market by 50%, according to ABB.
- Foxconn has begun testing the technology in consumer electronics assembly operations, with broader availability to 60,000 RobotStudio users expected in H2 2026.
- WORKR, a California robotics firm, will showcase the solution at NVIDIA GTC 2026, scheduled for March 16–19 in San Jose.
ABB Robotics revealed on Monday a collaboration with NVIDIA designed to solve a persistent challenge in industrial automation — achieving consistency between simulated robot behavior and real-world performance.
The Swiss robotics manufacturer will embed NVIDIA’s Omniverse libraries directly into its RobotStudio programming and simulation platform. This integration forms the foundation of RobotStudio HyperReality, scheduled for complete availability in the second half of 2026.
The collaboration addresses what engineers call the “sim-to-real” gap. Traditional simulations have historically failed to capture authentic factory conditions including lighting variations, shadow patterns, material textures, and physical inconsistencies. These limitations have required manufacturers to invest substantial time and resources reconciling simulated and actual environments.
ABB reports its solution achieves up to 99% accuracy in closing this gap. The company operates the sole virtual controller in the industry running identical firmware to its physical hardware, creating alignment between simulation outcomes and operational reality.
The company’s Absolute Accuracy technology further minimizes robot positioning errors from a typical 8–15mm range down to approximately 0.5mm, enabling precision applications such as electronics assembly.
Platform Capabilities
Manufacturers adopting RobotStudio HyperReality gain the ability to design, validate, and refine production lines in virtual environments before physical implementation. ABB projects this approach can reduce setup and commissioning times by up to 80%.
The platform also enables cost reductions of up to 40%, primarily by eliminating physical prototype requirements during development phases. Complex product time-to-market can be shortened by 50%, based on ABB’s internal assessments.
The system leverages synthetic data to prepare robots for diverse tasks and production configurations. Following virtual training, robots transition to production lines with the promised accuracy levels.
ABB is evaluating the integration of NVIDIA’s Jetson edge computing platform into its Omnicore controller, which would support real-time AI inference directly at the robot level.
Initial Implementation Partners
Foxconn, the global leader in electronics contract manufacturing, serves as the inaugural company testing the combined solution. The manufacturer is deploying RobotStudio HyperReality to prepare assembly robots for consumer electronics production — operations requiring precise pick-and-place movements across diverse device configurations.
Foxconn’s Chief Digital Officer, Dr. Zhe Shi, stated the accuracy and fidelity the platform provides “just wasn’t possible in simulation and digital twins” previously.
WORKR, a California-based robotic workforce provider, has also adopted the platform. During NVIDIA GTC 2026 in San Jose (March 16–19), WORKR will present AI-driven robotic systems powered by ABB technology that operate without programming expertise.
WORKR CEO Ken Macken described the partnership as focused on making industrial AI “deployable today,” with particular emphasis on small and medium manufacturers experiencing labor shortages.
ABB confirmed RobotStudio HyperReality will become accessible to all 60,000 existing RobotStudio customers upon its second half 2026 launch.
ABBN stock declined 4.22% and NVDA dropped 3.01% at the time of reporting.
