In the context of automated feeding (orienting) of industrial parts, we study the algorithmic design of traps in the bowl feeder track that filter out all but one orientation of a given polyhedral part. We propose a new class of traps that removes a V-shaped portion of the track. The proposed work advances the state-of-the-art in algorithmic trap design by extending earlier work [1, 6, 17]—which focuses solely on 2D parts— to 3D parts, and by incorporating a more realistic part motion model in the design algorithm. We exploit the geometric structure of the design problem and build on concepts and techniques from computational geometry to obtain an efficient algorithm that reports the complete set of valid traps.
Onno C. Goemans, A. Frank van der Stappen