Dr Andrew Colligan

Andrew completed his PhD in 2022 from Queen’s University Belfast after previously obtaining Master of Mechanical Engineering also from Queen’s. His PhD project titled ”Deep Learning for Boundary Representation CAD Models” focussed on using neural networks for learning from B-Rep CAD models specifically the task of automatic feature recognition. This involved working with a range of shape representations i.e., meshes, point clouds and voxels, as well as state-of-the-art 3D shape deep learning methodologies. The novelty involved the creation of a new hierarchical graph shape representation to encode the B-Rep and a novel hierarchical graph convolutional network for learning from the proposed shape representation. His research interests include 3D shape deep learning as well as deep learning for other engineering applications such as manufacturing. Andrew joined the RIED project in January 2022, where he has been working across threads on the integration of the group’s neuroevolution algorithm with certain shape and simulation information.