ROBOTURK: A Crowdsourcing Platform for Robotic Skill Learning through Imitation

By Ajay Mandlekar, Yuke Zhu, Animesh Garg, Jonathan Booher, Max Spero, Albert Tung, Julian Gao, John Emmons, Anchit Gupta, Emre Orbay and 2 more. 2018

387 citations recorded by the source corpus.

Imitation Learning has empowered recent advances in learning robotic manipulation tasks by addressing shortcomings of Reinforcement Learning such as exploration and reward specification. However, research in this area has been limited to modest-sized datasets due to the difficulty of collecting large quantities of task demonstrations through existing mechanisms. This work introduces RoboTurk to address this challenge. RoboTurk is a crowdsourcing platform for high quality 6-DoF trajectory based teleoperation through the use of widely available mobile devices (e.g. iPhone). We evaluate RoboTurk on three manipulation tasks of varying timescales (15-120s) and observe that our user interface is statistically similar to special purpose hardware such as virtual reality controllers in terms of task completion times. Furthermore, we observe that poor network conditions, such as low bandwidth and high delay links, do not substantially affect the remote users' ability to perform task demonstrations successfully on RoboTurk. Lastly, we demonstrate the efficacy of RoboTurk through the collection of a pilot dataset; using RoboTurk, we collected 137.5 hours of manipulation data from remote workers, amounting to over 2200 successful task demonstrations in 22 hours of total system usage. We show that the data obtained through RoboTurk enables policy learning on multi-step manipulation tasks with sparse rewards and that using larger quantities of demonstrations during policy learning provides benefits in terms of both learning consistency and final performance. For additional results, videos, and to download our pilot dataset, visit $\href{this http URL}{\texttt{roboturk.stanford.edu}}$

Subfields: Biomedical Engineering, Computer Science Applications, Control and Systems Engineering, Robotics & Control

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Biomedical Engineering

In Computer Science Applications

In Control and Systems Engineering

In Robotics & Control

Citations in the corpus

Learning to Generalize Across Long-Horizon Tasks from Human Demonstrations (2020, methodology): “We leverage the RoboTurk robot teleoperation interface [27, 28] to collect task demonstrations from humans.”
RoboNet: Large-Scale Multi-Robot Learning (2019, methodology): “eds of classes, the ImageNet Large-Scale Visual Recognition Challenge (ILSVRC) contained one thousand classes. Our work is inspired by this idea: while prior robotic manipulation methods and datasets [5,46,47, 26,48,7,49] generally consider a single robot at a time, our dataset includes 7 different ”
Generalization Through Hand-Eye Coordination: An Action Space for Learning Spatially-Invariant Visuomotor Control (2021, methodology): “The human demonstrations are recorded using RoboTurk [18, 20], which is a platform that allows users to demonstrate manipulation tasks through low-latency teleoperation with a smartphone interface.”
Learning Multi-Arm Manipulation Through Collaborative Teleoperation (2020, background): “smartphones, which are more ubiquitous and can provide the same quality of teleoperation as VR [27].”
Accelerating Reinforcement Learning with Learned Skill Priors (2020, background): “via human teleoperation [28, 9, 20, 19] or any combination of these.”
robosuite: A Modular Simulation Framework and Benchmark for Robot Learning (2020, methodology): “In particular, the SpaceMouse from 3Dconnexion is a 3D mouse that we used extensively for collecting demonstrations [17, 10] and debugging task designs.”
IRIS: Implicit Reinforcement without Interaction at Scale for Learning Control from Offline Robot Manipulation Data (2019, methodology): “3: Manipulation Results: We present a comparison of IRIS against several baselines on the Robosuite Lift and RoboTurk Cans datasets (left two plots).”
Meta-World: A Benchmark and Evaluation for Multi-Task and Meta Reinforcement Learning (2019, background): “e challenges in applying metalearning to robotic manipulation problems, including both parameteric and non-parametric variation in tasks. There is a long history of robotics benchmarks [28], datasets [29,30,31,32,33,34,35], competitions [36] and standardized object sets [37,38] that have played an i”
RLBench: The Robot Learning Benchmark & Learning Environment (2019, background): “For evaluating imitation learning systems in particular, RoboTurk [28] was a recent attempt to leverage crowd sourcing to obtain data for tasks, but because of this the system has only three tasks.”
Continual Learning for Robotics (2019, background): “, 2018) for robot manipulation and RoboTurk for robotic skill learning through imitation (Mandlekar et al., 2018), CRLMaze extension of VizDoom (Lomonaco et al.”
SAPIEN: A SimulAted Part-Based Interactive ENvironment (2020, background): “One direct way to address the problem is to train robots by interacting with the real environment [29, 4, 26].”
Bridge Data: Boosting Generalization of Robotic Skills with Cross-Domain Datasets (2021, methodology): “Most existing robot datasets, such as MIME [16], DAML [10], RoboTurk [17], [18], and many others [19], [20], [21], [22], [23], [24], [5] only feature a single domain, making them difﬁcult to use for boosting the generalization in other domains.”
Learning Interactive Real-World Simulators (2023, background): “…5], video captioning and question answering data contain rich high-level descriptions but little low-level movement detail [6, 7], human activity data contains rich human action but little mechanical motion [8, 9], and robotics data contains rich robot action but are limited in quantity [10, 11].”
Scaling data-driven robotics with reward sketching and batch reinforcement learning (2019, background): “Roboturk [47] collects crowd-sourced demonstrations for three tasks with the mobile platform.”
Recent Advances in Robot Learning from Demonstration (2020, background): “Additionally, access to remote demonstrators opens the opportunity for crowdsourcing demonstrations at a large scale (30, 31, 32, 33).”
iGibson 1.0: A Simulation Environment for Interactive Tasks in Large Realistic Scenes (2020, methodology): “Second, we collect demonstrations for imitation learning through continuous teleoperation using a system similar to Roboturk [67] for mobile-manipulation tasks.”
UniVLA: Learning to Act Anywhere with Task-centric Latent Actions (2025, neutral): “Fractal [11] 13.9% Kuka [36] 6.3% Bridge [78] 6.8% Taco Play [57] 3.5% Jaco Play [21] 0.6% Berkeley Cable Routing [51] 0.3% Roboturk [55] 2.8% Viola [99] 1.1% Berkeley Autolab UR5 [15] 1.4% Toto [95] 2.4% Language Table [53] 5.2% Stanford Hydra Dataset [7] 5.3% Austin Buds Dataset [98] 0.3% NYU…”