MotorFactory: A Blender Add-on for Large Dataset Generation of Small Electric Motors

To enable automatic disassembly of different product types with uncertain condition and degree wear in remanufacturing, agile production systems that can adapt dynamically to changing requirements are needed. Machine learning algorithms be employed due their generalization capabilities from various variants products. However, reality, datasets a diversity samples used train models difficult obtain the initial period. This may cause bad performances when system tries new unseen input data future. In order generate large for purposes, our project, we present Blender add-on named MotorFactory customized mesh motor instances. allows create which, complemented additional add-ons, further synthetic RGB images, depth normal segmentation ground truth masks 3D point cloud point-wise semantic labels. The created tasks including type classification, object detection decentralized material transfer tasks, part handling or even reinforcement learning-based robotics control view-planning.