30 March 2026 | News
Image Courtesy: Public Domain
Universal Robots (UR) unveiled its UR AI Trainer this week at GTC 2026. Developed in collaboration with Scale AI , the AI Trainer symbolizes a shift in robotics from pre-written applications to fully AI-driven operating systems.
“Our customers, from large enterprises to AI research labs, are no longer just asking for AI capabilities,” said Anders Beck, Vice President of AI Robotics Products at Universal Robots. “They need a way to collect high-precision, synchronized robotic and visual data to train AI models on the same robot that will be deployed. Our AI Trainer is the industry’s first AI model training solution that goes directly from the lab to the factory.”
Achieving AI- Ready Data Acquisition
AI robot training is often hampered by incomplete hardware and low-precision data acquisition. Current training data is collected using research robots unsuitable for production environments, and most systems rely solely on visual feedback, making tasks requiring fine motor skills or extensive contact extremely difficult. Beck stated, "AI Trainer directly addresses these obstacles."
"By leveraging our unique direct torque control and force feedback capabilities, and training with the same powerful hardware across more than 100,000 industrial deployment sites, we enable developers to directly influence how robots actually interact with their surroundings."
The partnership with Scale AI enables a flywheel effect of integrated robotic data.
UR AI Trainer enables human operators to guide UR robots through tasks in a leader-follower setup, capture high-quality, synchronous multimodal data during live demonstrations, and build the structured datasets needed to train vision-language-action (VLA). Executed on UR's AI Accelerator platform, AI Trainer combines collaborative industrial robots with Scale AI software to enable scalable data capture in production environments, supporting continuous optimization of physical AI systems.
“Universal Robots is a leader in industrial robotics, and its global footprint provides an ideal foundation for data acquisition and AI deployment,” said Ben Levin, General Manager of Entity AI at Scale AI. “Together, we’ve built an integrated robotic data flywheel that enables customers to train, deploy, and improve their AI models at unprecedented speeds.”
UR and Scale AI will release a large-scale industrial dataset later this year, collected using UR robots.
Experience AI Trainer at GTC
Visitors to UR's GTC booth can guide two UR3e "Leader" robots and control two UR7e "Follower" robots via haptic input. This device allows visitors to perform advanced smartphone packaging tasks through haptic feedback for imitation learning and VLA training. The demonstration data is recorded in real-time on Scale's platform and can be replayed directly on the AI Trainer.
