AutoDex: An Automated Real-World System for Dexterous Grasping Data Collection
2026-06-22 • Robotics
RoboticsMachine Learning
AI summaryⓘ
The authors designed AutoDex, a system that collects real-world data on how well robot hands can grasp objects without needing humans to help at each step. AutoDex uses many cameras to see objects even when parts are hidden, moves the robot carefully, checks if the grasp worked, and resets objects for new tries automatically. This makes it faster and more reliable than doing the same work by hand or just in simulation. They tested it with different robot hands and objects, showing it can make a large, helpful database for training better robot grasping.
dexterous graspingrobot perceptionautomationrobot manipulationmulti-view camerasdata collectionrobot grasp evaluationteleoperationsimulation validationrobot reset mechanisms
Authors
Mingi Choi, Gunhee Kim, Jisoo Kim, Taeksoo Kim, Taeyun Ha, Jongbin Lim, Hanbyul Joo
Abstract
Learning robust dexterous grasping requires real-world data that records the physical outcomes of grasp attempts. Such data is hard to obtain at scale: teleoperation yields valid physical outcomes but is slow and operator-biased, while simulation-based generation is cheap and scalable but cannot certify contact validity. A natural solution is to generate candidate grasps and verify them on real hardware, but this scales only if the entire collection loop (perception, execution, labeling, and reset) runs without human intervention. We present AutoDex, an automated real-world data-collection system that closes this loop: for each candidate from a replaceable generator, it localizes the object under severe hand-object occlusion with dense 20-camera perception, executes collision-monitored robot motions, labels lift-and-hold success or failure, and actively resets the object between trials to expose additional candidates across stable poses. The result is a reusable database of physically labeled grasp trials that downstream systems can query by retrieval and feasibility filtering. Using AutoDex, we collect 3,593 grasp trials across Allegro and Inspire hands on 100 diverse objects, with synchronized multi-view observations and robot-state logs. For a matched 500-trajectory collection, AutoDex requires 10.3 h versus 49.4 h for teleoperation, yielding a 4.8x throughput improvement, and grasps retrieved from the AutoDex-validated database succeed 76% versus 34% for simulation-only validation. Code and data will be publicly released.