Opening Articulated Structures in the Real World
Arjun Gupta, Michelle Zhang, Rishik Sathua, Saurabh Gupta

TL;DR
This paper investigates the challenges of enabling mobile robots to open unseen articulated objects in real-world settings, revealing that modular systems outperform end-to-end learning and perception is the main bottleneck.
Contribution
It presents a comprehensive system for opening articulated structures in real environments and provides large-scale empirical insights into system design and perception challenges.
Findings
Modular systems outperform end-to-end learned systems.
Perception is the primary bottleneck, not end-effector control.
Articulation parameter estimation models struggle with robot-centric viewpoints.
Abstract
What does it take to build mobile manipulation systems that can competently operate on previously unseen objects in previously unseen environments? This work answers this question using opening of articulated structures as a mobile manipulation testbed. Specifically, our focus is on the end-to-end performance on this task without any privileged information, i.e. the robot starts at a location with the novel target articulated object in view, and has to approach the object and successfully open it. We first develop a system for this task, and then conduct 100+ end-to-end system tests across 13 real world test sites. Our large-scale study reveals a number of surprising findings: a) modular systems outperform end-to-end learned systems for this task, even when the end-to-end learned systems are trained on 1000+ demonstrations, b) perception, and not precise end-effector control, is the…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAdvanced Manufacturing and Logistics Optimization
MethodsFocus
