“Physical AI fundamentally changes what robots can do in industrial applications," said Craig McDonnell, Managing Director, Business Line Industries, ABB Robotics. "Some of the greatest opportunities for physical AI are in applications that have traditionally been difficult to automate because of high variability. Fashion Inductor with AI Robot Trainer is our first end-to-end vision-based physical AI solution, demonstrating how intelligent automation can be deployed, operated and continuously optimized at industrial scale, helping system integrators accelerate deployment, reduce project risk and maintain performance over time.”
Designed to handle the high variability of fashion and apparel logistics, where products, packaging, and operating conditions constantly change, Fashion Inductor delivers high-performance sorting in real-world logistics environments. The system achieves throughput of up to 1,300 items per hour per cell and targets picking accuracy above 99.5%, supported by ABB Robotics’ AI-powered Automatic Path Planning, precise motion control, and application-specific vision technology.
The integration of AI Robot Trainer introduces vision-based physical AI capabilities that reduce engineering and commissioning efforts by up to 75%. The solution provides pre-trained AI vision models tailored to fashion applications, specialized tools, and rapid retraining capabilities, enabling system integrators to deploy, adapt, and continuously optimize applications as requirements change without requiring specialized AI expertise.
The upgraded Fashion Inductor also features a new digital twin that precisely replicates the module’s software environment. It connects the Application Controller Platform, Automatic Path Planning, AI Robot Trainer, and RobotStudio®. This integration streamlines commissioning and provides greater setup flexibility, including support for third-party components such as grippers.
AI models can be developed and continuously refined throughout the system lifecycle to meet actual performance targets and adapt to changing operating conditions. AI Robot Trainer supports structured, rapid retraining tailored to individual cells through a continuous closed-loop workflow. This process monitors robot performance, detects anomalies, generates new training data, and deploys model improvements back to the physical robot.
This lifecycle approach is built on ABB Robotics’ core platforms. RobotStudio serves as the foundation for virtual design, simulation, and validation, while the OmniCore™ robot controller executes operations in production environments.
Fashion Inductor represents the first step in ABB Robotics’ broader introduction of vision-based physical AI, a key component of its development of more autonomous and versatile robotics under the Autonomous Versatile Robotics (AVR™) initiative. The company plans to extend the AI Robot Trainer-powered approach to additional vision-based applications through the first quarter of 2027, bringing adaptive intelligence and continuous lifecycle management to the next generation of increasingly autonomous robotic systems.
