Robust 2D Assembly Sequencing via Geometric Planning with Learned Scores
This addresses the challenge of efficient and robust assembly sequencing for robotics, though it is incremental as it builds on existing geometric planning and simulation methods.
The paper tackled the problem of computing robust 2D assembly plans by combining geometric planning with a deep neural network trained to predict robustness scores, resulting in an order of magnitude faster planning than physics simulation.
To compute robust 2D assembly plans, we present an approach that combines geometric planning with a deep neural network. We train the network using the Box2D physics simulator with added stochastic noise to yield robustness scores--the success probabilities of planned assembly motions. As running a simulation for every assembly motion is impractical, we train a convolutional neural network to map assembly operations, given as an image pair of the subassemblies before and after they are mated, to a robustness score. The neural network prediction is used within a planner to quickly prune out motions that are not robust. We demonstrate this approach on two-handed planar assemblies, where the motions are one-step translations. Results suggest that the neural network can learn robustness to plan robust sequences an order of magnitude faster than physics simulation.