• Skip to primary navigation
  • Skip to main content
  • Skip to primary sidebar
  • Skip to secondary sidebar
  • Skip to footer
  • Home
  • Subscribe
  • Your Membership
    • Edit Your Profile
  • Services
    • Advertising
    • Case studies
    • Design
    • Email marketing
    • Lead generation
    • Magazine
    • Press releases
    • Publishing
    • Sponsored posts
    • Webcasting
    • Webinars
    • White papers
    • Writing
  • Shop
    • My Account
    • Cart
  • About
    • Contact
    • Privacy
    • Terms of use
  • Events

Robotics & Automation News

Market trends and business perspectives

  • News
  • Features
  • Video
  • Webinars
  • White papers
  • Press releases
  • Featured companies
    • AMD Xilinx
    • BlueBotics
    • Elite Robot
    • RGo Robotics
    • SICK Sensor Intelligence
    • Vicor Power

Nvidia unveils new reinforcement learning research at ICRA 2019

May 31, 2019 by Sam Francis

Nvidia researchers from the newly opened robotics research lab in Seattle, Washington have presented a new proof of concept reinforcement learning approach that aims to enhance how robots trained in simulation will perform in the real world. (See video below.)

The work was presented at the recent International Conference on Robotics and Automation (ICRA) in Montreal, Canada.

The research is part of a growing trend in the deep learning and robotics community that relies on simulation for training.

Since the method is virtual, there’s no risk of damage or injury, allowing the robot to train for potentially an unlimited number of times, before deploying to the real world.

One way to describe simulation training is to compare it to how astronauts train on earth for key space missions.

They learn to withstand the punishing G-forces that come with space travel, rehearse and practice all aspects of a mission, and how to perform critical operations so they can per form them flawlessly in space.

Reinforcement learning in simulation aims to do the same but with robots.

Ankur Handa, one of the lead researchers on the project, says: “In robotics, you generally want to train things in simulation because you can cover a wide spectrum of scenarios that are difficult to get data for in the real world.

“The idea behind this work is to train the robot to do something in the simulator that would be tedious, and time-consuming in real life.”

Handa says that one of the challenges researchers in the reinforcement learning robotics community face is the discrepancy between what is in the real world and simulator.

“Due to the imprecise simulation models and lack of high fidelity replication of real-world scenes, policies learned in simulations often cannot be directly applied on real-world systems, a phenomenon that is also known as the reality gap,” the researchers state in their paper.

“In this work, we focus on closing the reality gap by learning policies on distributions of simulated scenarios that are optimized for a better policy transfer.”

“Rather than manually tuning the randomization of simulations, we adapt the simulation parameter distribution using a few real world roll-outs interleaved with policy training,” Handa says. “We’re essentially creating a replica of the real world in the simulator.”

Using a cluster of 64 Nvidia Tesla V100 GPUs, with the cuDNN-accelerated TensorFlow deep learning framework, the researchers trained a robot to perform two tasks: placing a peg in a hole and opening a drawer.

For the simulation, the team used the Nvidia FleX physics engine to simulate and develop the SimOpt algorithm described in this research work.

For both tasks, the robot learns from over 9600 simulations each over around 1.5-2 hours, allowing it to swing a peg into a hole and open a drawer accurately.

“Closing the simulation to reality transfer loop is an important component for a robust transfer of robotic policies,” the researchers say.

“In this work, we demonstrated that adapting simulation randomization using real world data can help in learning simulation parameter distributions that are particularly suited for a successful policy transfer without the need for exact replication of the real world environment.”

Print Friendly, PDF & Email

Share this:

  • Print
  • Facebook
  • LinkedIn
  • Reddit
  • Twitter
  • Tumblr
  • Pinterest
  • Skype
  • WhatsApp
  • Telegram
  • Pocket

You might also like…

Filed Under: Computing, Features Tagged With: aims, challenges, community, data, gap, handa, icra, learning, nvidia, perform, policies, presented, real, real-world, reality, reinforcement, researchers, robot, robotics, robots, scenarios, simulation, simulator, space, train, training, work

Join the Robotics & Automation News community

Primary Sidebar

Latest articles

  • ASI and SICK optimize and automate logistics truck yard operations
  • Fugro provides uncrewed surface vessel to TAQA
  • Radial selects Covariant to automate e-commerce fulfillment with AI-powered robotics
  • Aerones demonstrates ‘first’ robot for wind turbine maintenance and repair
  • Electric Future: Two-wheels good, four wheels not so much
  • Electric Future: Battery production facilities bloom on both sides of the Atlantic
  • Electric Future: MIT showcases electric autonomous boat technology
  • Electric Future: Siemens providing software platform to support nascent electric airplane market
  • Electric Future: Stalled car revolution
  • Scythe Robotics secures $42 million new financing to accelerate production of zero-emissions autonomous mower

Most Read

  • Top 20 electric vehicle charging station companies
    Top 20 electric vehicle charging station companies
  • Why is My Car Key Stuck in the Ignition?
    Why is My Car Key Stuck in the Ignition?
  • Berkshire Grey and Locus Robotics combine to offer ‘industry-first’ cross-platform robotic automation
    Berkshire Grey and Locus Robotics combine to offer ‘industry-first’ cross-platform robotic automation
  • Mercedes-Benz becomes ‘world’s first’ automotive company to certify SAE Level 3 system for US market
    Mercedes-Benz becomes ‘world’s first’ automotive company to certify SAE Level 3 system for US market
  • Difference Between Three-Phase and Single-Phase Power
    Difference Between Three-Phase and Single-Phase Power
  • Universal Robots reports record annual revenue of $326 million
    Universal Robots reports record annual revenue of $326 million
  • Lockheed Martin Ventures invests in on-demand manufacturing startup Machina Labs
    Lockheed Martin Ventures invests in on-demand manufacturing startup Machina Labs
  • Scientists have found more water in space than they ever knew possible
    Scientists have found more water in space than they ever knew possible
  • The Best Mechanical Engineering Design Software in 2022
    The Best Mechanical Engineering Design Software in 2022
  • Electric Future: Siemens providing software platform to support nascent electric airplane market
    Electric Future: Siemens providing software platform to support nascent electric airplane market

Overused words

ai applications automated automation automotive autonomous business china companies company control customers data design development digital electric global industrial industry logistics machine manufacturing market mobile operations platform process production robot robotic robotics robots safety software solution solutions system systems technologies technology time vehicle vehicles warehouse

Secondary Sidebar

Latest news

  • ASI and SICK optimize and automate logistics truck yard operations
  • Fugro provides uncrewed surface vessel to TAQA
  • Radial selects Covariant to automate e-commerce fulfillment with AI-powered robotics
  • Aerones demonstrates ‘first’ robot for wind turbine maintenance and repair
  • Electric Future: Two-wheels good, four wheels not so much
  • Electric Future: Battery production facilities bloom on both sides of the Atlantic
  • Electric Future: MIT showcases electric autonomous boat technology
  • Electric Future: Siemens providing software platform to support nascent electric airplane market
  • Electric Future: Stalled car revolution
  • Scythe Robotics secures $42 million new financing to accelerate production of zero-emissions autonomous mower

Footer

We are…

Robotics and Automation News was established in May, 2015, and is now one of the most widely-read websites in its category.

Please consider supporting us by becoming a paying subscriber, or through advertising and sponsorships, or by purchasing products and services through our shop – or a combination of all of the above.

Thank you.

Independent

Archivists

May 2019
M T W T F S S
 12345
6789101112
13141516171819
20212223242526
2728293031  
« Apr   Jun »

Complex

Old-skool

This website and its associated magazine, and weekly newsletter, are all produced by a small team of experienced journalists and media professionals.

If you have any suggestions or comments, feel free to contact us at any of the email addresses on our contact page.

We’d be happy to hear from you, and will always reply as soon as possible.

Future-facing

Free, fair and legal

We support the principles of net neutrality and equal opportunities.

Member of The Internet Defense League

Copyright © 2023 · News Pro on Genesis Framework · WordPress · Log in

We use cookies on our website to give you the most relevant experience by remembering your preferences and repeat visits. By clicking “Accept”, you consent to the use of ALL the cookies.
Do not sell my personal information.
Cookie SettingsAccept
Manage consent

Privacy Overview

This website uses cookies to improve your experience while you navigate through the website. Out of these, the cookies that are categorized as necessary are stored on your browser as they are essential for the working of basic functionalities of the website. We also use third-party cookies that help us analyze and understand how you use this website. These cookies will be stored in your browser only with your consent. You also have the option to opt-out of these cookies. But opting out of some of these cookies may affect your browsing experience.
Necessary
Always Enabled
Necessary cookies are absolutely essential for the website to function properly. These cookies ensure basic functionalities and security features of the website, anonymously.
CookieDurationDescription
cookielawinfo-checkbox-analytics11 monthsThis cookie is set by GDPR Cookie Consent plugin. The cookie is used to store the user consent for the cookies in the category "Analytics".
cookielawinfo-checkbox-functional11 monthsThe cookie is set by GDPR cookie consent to record the user consent for the cookies in the category "Functional".
cookielawinfo-checkbox-necessary11 monthsThis cookie is set by GDPR Cookie Consent plugin. The cookies is used to store the user consent for the cookies in the category "Necessary".
cookielawinfo-checkbox-others11 monthsThis cookie is set by GDPR Cookie Consent plugin. The cookie is used to store the user consent for the cookies in the category "Other.
cookielawinfo-checkbox-performance11 monthsThis cookie is set by GDPR Cookie Consent plugin. The cookie is used to store the user consent for the cookies in the category "Performance".
viewed_cookie_policy11 monthsThe cookie is set by the GDPR Cookie Consent plugin and is used to store whether or not user has consented to the use of cookies. It does not store any personal data.
Functional
Functional cookies help to perform certain functionalities like sharing the content of the website on social media platforms, collect feedbacks, and other third-party features.
Performance
Performance cookies are used to understand and analyze the key performance indexes of the website which helps in delivering a better user experience for the visitors.
Analytics
Analytical cookies are used to understand how visitors interact with the website. These cookies help provide information on metrics the number of visitors, bounce rate, traffic source, etc.
Advertisement
Advertisement cookies are used to provide visitors with relevant ads and marketing campaigns. These cookies track visitors across websites and collect information to provide customized ads.
Others
Other uncategorized cookies are those that are being analyzed and have not been classified into a category as yet.
SAVE & ACCEPT