Part Identification and Localization via Deep Neural Networks

Symbio Robotics is developing a robust, fast, and low-cost perception engine for identifying parts and detecting their six degree of freedom pose. Our perception engine is driven by deep convolutional neural networks and draws upon recent advances in computer vision and deep learning. Current automation systems largely run blind, without perception driven feedback. Our proposed system seeks to close that feedback loop and unlock new levels of performance in industrial automation at this time of rapid change in the domain of manufacturing.