A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Polyhedron-bounded collision checks for robotic assembly of structural components
Abstract For existing robotic assembly studies in construction, the axis-aligned bounding box algorithm used is not capable enough to provide the exact approximation of rotated and irregularly-shaped structural components' volumetric extents. This could lead to mistakenly assumed collisions. This research bridges the gap by developing a robotic approach that can mathematically derive polyhedron-bounded boxes that provide a more accurate approximation of rotated and irregularly-shaped structural components' volumetric extents for collision checks. In the approach, structural components' polyhedron meshes are represented by vertexes and normal in the algebraic format and are compared in real time to identify touching faces. Testing of the approach reveals that it has satisfactory collision avoidance performance, where its generated outputs can be used to control the joints of a robotic manipulator for detouring obstacles. The strength of our approach is in exploiting entities' polyhedral features, where the constraints for testing collisions are set with objects' exact boundaries.
Highlights A polyhedron-bounded collision detection approach is developed for robotic assembly. Polyhedral algebraization is used to capture rotated and irregularly-shaped structural components' volumetric extents. Components' volumetric extents are represented by vertex and normal in the algebraic format. Components' vertex and normal matrices are compared in real time to identify collisions.
Polyhedron-bounded collision checks for robotic assembly of structural components
Abstract For existing robotic assembly studies in construction, the axis-aligned bounding box algorithm used is not capable enough to provide the exact approximation of rotated and irregularly-shaped structural components' volumetric extents. This could lead to mistakenly assumed collisions. This research bridges the gap by developing a robotic approach that can mathematically derive polyhedron-bounded boxes that provide a more accurate approximation of rotated and irregularly-shaped structural components' volumetric extents for collision checks. In the approach, structural components' polyhedron meshes are represented by vertexes and normal in the algebraic format and are compared in real time to identify touching faces. Testing of the approach reveals that it has satisfactory collision avoidance performance, where its generated outputs can be used to control the joints of a robotic manipulator for detouring obstacles. The strength of our approach is in exploiting entities' polyhedral features, where the constraints for testing collisions are set with objects' exact boundaries.
Highlights A polyhedron-bounded collision detection approach is developed for robotic assembly. Polyhedral algebraization is used to capture rotated and irregularly-shaped structural components' volumetric extents. Components' volumetric extents are represented by vertex and normal in the algebraic format. Components' vertex and normal matrices are compared in real time to identify collisions.
Polyhedron-bounded collision checks for robotic assembly of structural components
Gao, Yifan (author) / Shu, Jiangpeng (author) / Xiao, Wenkai (author) / Jin, Zhenfen (author)
2023-04-22
Article (Journal)
Electronic Resource
English
| TIBKAT | 1988
Structural Checks for Decennial Insurance
| British Library Conference Proceedings | 1993
Quick Checks for Structural Computer Analysis
| British Library Conference Proceedings | 1994
Habitable Polyhedron, Bogota, Colombia
| British Library Online Contents | 2011