For instance, Siemens’ SIMATIC Robot Pick AI expands on this vision of adaptability, transforming standard industrial robots—once limited to rigid, repetitive tasks—into complex machines. Trained on synthetic data—virtual simulations of shapes, materials, and environments—the AI prepares robots to handle unpredictable tasks, like picking unknown items from chaotic bins, with over 98% accuracy. When mistakes occur, the system learns, improving through real-world feedback. Crucially, this isn’t only a one-robot fix. Software updates scale across entire fleets, upgrading robots to work more flexibly and meet the rising demand for adaptive production.
One other example is the robotics firm ANYbotics, which generates 3D models of business environments that function as digital twins of real environments. Operational data, resembling temperature, pressure, and flow rates, are integrated to create virtual replicas of physical facilities where robots can train. An energy plant, for instance, can use its site plans to generate simulations of inspection tasks it needs robots to perform in its facilities. This speeds the robots’ training and deployment, allowing them to perform successfully with minimal on-site setup.
Simulation also allows for the near-costless multiplication of robots for training. “In simulation, we will create hundreds of virtual robots to practice tasks and optimize their behavior. This enables us to speed up training time and share knowledge between robots,” says Péter Fankhauser, CEO and co-founder of ANYbotics.
Because robots need to know their environment no matter orientation or lighting, ANYbotics and partner Digica created a way of generating hundreds of synthetic images for robot training. By removing the painstaking work of collecting huge numbers of real images from the shop floor, the time needed to show robots what they should know is drastically reduced.
Similarly, Siemens leverages synthetic data to generate simulated environments to coach and validate AI models digitally before deployment into physical products. “Through the use of synthetic data, we create variations in object orientation, lighting, and other aspects to make sure the AI adapts well across different conditions,” says Vincenzo De Paola, project lead at Siemens. “We simulate every little thing from how the pieces are oriented to lighting conditions and shadows. This enables the model to coach under diverse scenarios, improving its ability to adapt and respond accurately in the true world.”
Digital twins and artificial data have proven powerful antidotes to data scarcity and expensive robot training. Robots that train in artificial environments could be prepared quickly and inexpensively for wide varieties of visual possibilities and scenarios they could encounter in the true world. “We validate our models on this simulated environment before deploying them physically,” says De Paola. “This approach allows us to discover any potential issues early and refine the model with minimal cost and time.”
This technology’s impact can extend beyond initial robot training. If the robot’s real-world performance data is used to update its digital twin and analyze potential optimizations, it may well create a dynamic cycle of improvement to systematically enhance the robot’s learning, capabilities, and performance over time.
The well-educated robot at work
With AI and simulation powering a brand new era in robot training, organizations will reap the advantages. Digital twins allow corporations to deploy advanced robotics with dramatically reduced setup times, and the improved adaptability of AI-powered vision systems makes it easier for corporations to change product lines in response to changing market demands.
