Dex-Net as a Service (DNaaS): A Cloud-Based Robust Robot Grasp Planning System

Abstract

Accessing software resources via the Cloud has become increasingly popular as a means to configure and manage automation systems with reduced infrastructure overhead. Dex-Net as a Service (DNaaS) is a cloud-based grasp planning system for parallel-jaw grippers that provides a graphical user interface and API access to Dex-Net 1.0, a robust grasp planning system based on wrench mechanics and stochastic sampling. DNaaS allows anyone online to compute parallel-jaw grasps on triangular meshes and visualize the results. This paper presents system architecture, examples of generated grasps, and timing data for DNaaS. DNaaS takes under 75 seconds on average to process grasp requests. DNaaS is available at http://automation.berkeley.edu/dex-net.