Collision avoidance and adaptive path planning in machine tools by matching live image data with a geometric simulation

A major cause for unplanned downtime in small-series machining are collisions. While there are solutions to avoid collisions using geometric simulations, these do not cover collisions caused by setup errors. To address this problem from batch size one, a system has been developed which matches a geometric simulation with image data to detect deviations, modify the simulation and recalculate NC-Code to fit reality. Building on previous work regarding image-preprocessing, an iterative matching algorithm is developed, as well as a microservice based system-architecture which allows the integration of matching, adaptive path planning and collision avoidance simulation. The system is validated on a machining center.