3D Vision, AI Power Real-Time Robot Guidance, Enabling Tasks With Moving Parts

Inbolt turned heads at Universal Robots’ Collaborate 2025 event, held January 27 in Novi, MI, with its real-time industrial cobot guidance that leverages 3D vision and artificial intelligence (AI) to guide robots and cobots to handle parts as they move, or perform tasks on moving targets. 

Inbolt’s setup comprises a 3D camera mounted on the robot, Inbolt’s real-time workpiece localization AI and GuideNow Studio, a web application for the training Inbolt’s AI algorithms for each CAD model or 3D scan. Once a scan is made, the system can follow desired location points, no matter the orientation. AI continually refines locating by assessing various orientations and incorporating those orientations and extrapolating other orientations. Inbolt has robot applications in the field for screw insertion, pick-and-place, inspection and other tasks. This technology promises a solution for handling parts and components as they travel, ideal for assembly-line work, part transfer from conveyors and more, only limited by the speed of the robot and processing operation. 

Its AI-powered system quickly processes large amounts of 3D data and determines the position and orientation of a workpiece, then adapts a robot's trajectory in real time, according to Albane Dersy, co-founder and COO of Inbolt, who demonstrated the technology for MetalForming.

The company’s proprietary software reportedly is 100 times faster than leading algorithms, and can work with any hardware. The solution can reduce automation costs and improve automation flexibility by functioning in all types of environments: structured, semi-structured and unstructured. And, according to Inbolt officials, real-time robot guidance makes it possible to react dynamically to unplanned events.