Hardware-agnostic machine vision for real-time robotic autonomy

ABB Robotics has launched OmniCore EyeMotion, a software solution that enables industrial robots to process spatial data using third-party cameras and sensors for real-time environmental adaptation. The technology provides machine vision capabilities for automated systems deployed in the automotive, electronics, logistics, packaging, and food and beverage sectors.

Hardware-agnostic perception and rapid deployment
The integration of machine vision in industrial settings has historically relied on specialized optical setups to process three-dimensional spatial data. The new software architecture bypasses this limitation by allowing industrial control systems to process 2D and 3D vision data from standard third-party sensors. Utilizing a web-based interface, engineers can configure connected cameras via drag-and-drop programming. The system integrates directly with the RobotStudio simulation environment, a mechanism that reduces total commissioning time by up to 90 percent compared to custom-coded vision integration protocols.

Real-time path planning and operational applications
Targeting tasks such as automated quality control and item sorting, the vision software continuously captures and processes environmental data to adjust robotic movements. For highly variable environments, the platform functions in conjunction with continuous path-planning algorithms. This combination allows the robotic controller to dynamically calculate and execute collision-free trajectories around moving objects or sudden obstacles without requiring manual intervention. By optimizing these spatial calculations during operation, the system can decrease operational cycle times by up to 50 percent.




Advancing autonomous versatile robotics
The development of hardware-independent visual reasoning aligns with broader industry shifts toward autonomous manufacturing systems. "Over the past decade, we’ve advanced artificial intelligence (AI)-based vision to give robots ‘eyes’ capable of perceiving the world in 3D, but until now this required advanced camera hardware," says Marc Segura, President of ABB Robotics. The software serves as a foundational component for next-generation automated architectures, where continuous sensor feedback and generative artificial intelligence will manage complex tasks requiring precision, speed, and spatial mobility.

Additional Context
This section details technical specifications and competitive benchmarking not included in the original news release.

Within the industrial machine vision sector, hardware-agnostic robotic perception software operates in the same technical category as platforms developed by manufacturers such as Fanuc and Universal Robots. Benchmarking in this software class relies on measurable parameters including image processing latency, sensor compatibility range, and the reduction in programming hours required for deployment. The decoupling of vision processing software from proprietary camera hardware provides a specific operational standard for evaluating integration speed and system modularity against conventional robotics ecosystems that require vertically integrated optical components.

Edited by Aishwarya Mambet, Induportals Editor, with AI assistance.

www.abb.com