Collision Prediction Framework Using 3D Geometrical Digital Twins and Smart Manufacturing Middleware
Year: 2025
Type: article
Abstract: This paper presents a framework for predicting and preventing machine collisions using key concepts of smart manufacturing, including middleware and digital twins. The proposed approach integrates trajectory prediction, collision detection in 3D geometrical digital twins, and real-time feedback to machines. A digital twin interfaced with MTConnect middleware was used to reconstruct 3D geometries of the shop floor machine, and the Oriented Bounding Box (OBB) algorithm was employed to detect potential collisions. Future machine trajectories were predicted using a cross-correlation method that could compare the current trajectory with a reference motion cycle. By simulating these trajectories within the digital twin environment, potential collision points were identified in advance, enabling proactive avoidance strategies. The framework was validated through experimental scenarios, demonstrating its ability to accurately predict collision events and mitigate risks. Additionally, this approach can lead to enhanced fail-safe environments, streamlined production lines, and boosted productivity in smart factories by eliminating downtime caused by collisions and improving overall operational reliability.
Authors: Soon-Hong Hwang, Martin Byung-Guk Jun, Byung-Kwon Min, Huitaek Yun
Language: English
Subfield: Industrial and Manufacturing Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: bronze
Award: G4398737237
OpenAlex ID: W4412062204