ABB Robotics brings industrial-grade physical AI to fashion logistics

ABB Robotics launched its first end-to-end vision-based industrial-grade physical AI offering, integrating AI Robot Trainer into its Fashion Inductor module.

The company said that it can help system integrators deploy, scale and continuously optimize AI-powered fashion logistics applications with significantly reduced effort and risk.

High-performance sorting in fashion logistics

ABB Robotics said that Fashion Inductor is designed to meet the extreme variability of fashion and apparel logistics, where products, packaging and operating conditions change continuously. The company said that it is already delivering high-performance sorting for real-world logistics environments, with throughput of up to 1,300 items per hour per cell and picking accuracy targets above 99.5%. ABB Robotics said that this is enabled by the company’s AI-powered Automatic Path Planning, motion accuracy and application-specific vision technology.

“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.”

ABB Robotics said that the addition of AI Robot Trainer introduces vision-based physical AI capabilities that reduce engineering and commissioning effort by up to 75%, providing pre-trained, fashion-specific AI vision models, application-specific tools and rapid retraining capabilities that help system integrators deploy, adapt and continuously optimize applications as requirements evolve, without specialist AI expertise.

The company said that the upgraded Fashion Inductor also introduces a new digital twin, a replica of the module software, connecting Application Controller Platform, Automatic Path Planning, AI Robot Trainer and RobotStudio. ABB Robotics said that this accelerates commissioning and gives greater flexibility in set-up, including the ability to integrate third-party components such as grippers.

The company said that not only are AI models developed, but they are continually improved throughout the system lifecycle, against real performance targets and changing operating conditions. ABB Robotics said that AI Robot Trainer enables structured, rapid cell‑specific retraining through a continuous closed-loop workflow that monitors robot performance, identifies anomalies, generates new model training data and deploys improvements back to the physical robot.

The lifecycle is anchored in ABB Robotics’ core platforms, with RobotStudio providing the foundation for virtual design, simulation and validation, and the OmniCore robot controller serving as the execution layer in production.

ABB Robotics added that Fashion Inductor marks the first step in the company’s broader rollout of vision-based physical AI, a core part of developing more autonomous and versatile robotics (AVR). The company said that the same AI Robot Trainer-powered approach will continue to expand across additional ABB Robotics vision-based applications through Q1 2027, bringing adaptive, lifecycle-managed intelligence to the next generation of increasingly autonomous robotic systems.