---
layout: default
overview: true
---
<style>
.logos {
border-radius: 25px;
text-align: center;
margin-top: 0.5em;
-webkit-box-shadow: 0 3px 10px rgba(0,0,0,0.1);
-moz-box-shadow: 0 3px 10px rgba(0,0,0,0.1);
box-shadow: 0 3px 10px rgba(0,0,0,0.1);
}
.logos img {
  display: block;
  max-width:230px;
  max-height:45px;
  width: auto;
  height: auto;
  display: inline-block;
  padding-left: 0.9em;
padding-right: 0.9em;
margin-top: 0.3em;
margin-bottom: 0.3em;
}</style>
<section class="intro">
  <div class="grid">
    <div class="unit one-third center-on-mobiles" style="padding-top: 1em">
      <img src="https://s3.amazonaws.com/osim-rl/videos/running.gif" />
    </div>
    <div class="unit two-thirds center-on-mobiles" style="padding-top: 1em; padding-left: 1em">
      <p class="first">Reinforcement learning with musculoskeletal models in OpenSim</p>
    </div>
  </div>
</section>
<section class="features">
  <div class="grid">
    <div class="unit golden-large">
      <h2>NeurIPS 2019: Learn to Move - Walk Around</h2>
      <p>
  Design artificial intelligent controllers for the human body to accomplish diverse locomotion tasks. Participate in the NeurIPS 2019 challenge to win prizes and fame.
      </p>
      <a href="/docs/nips2019/">Learn more about the challenge &rarr;</a>
    </div>
    <div class="unit golden-small">
      <h2>OpenSim RL</h2>
      <p>
        Use our musculoskeletal reinforcement learning environment for other projects in computer science, neuroscience, biomechanics, etc.
      </p>
      <a href="/docs/home/">Learn more about osim-rl &rarr;</a>
    </div>
    <!--div class="clear"></div>
    <div class="unit whole center-on-mobiles logos" style="background-color: #ffffff;">
      <a href="https://news.stanford.edu/2017/08/07/virtual-competitors-vie-different-kind-athletic-title/"><img src="https://s3.amazonaws.com/osim-rl/logos/stanford.png" /></a>
      <a href="https://people.eecs.berkeley.edu/"><img src="https://s3.amazonaws.com/osim-rl/logos/berkeley.png" /></a>
      <a href="https://actu.epfl.ch/news/crowdai-the-open-data-science-challenge-platform/"><img src="https://s3.amazonaws.com/osim-rl/logos/epfl.png" /></a>
      <a href="http://mobilize.stanford.edu/"><img src="https://s3.amazonaws.com/osim-rl/logos/mobilize.png" /></a>
      <a href="https://www.crowdai.org/challenges/nips-2018-ai-for-prosthetics-challenge"><img src="https://s3.amazonaws.com/osim-rl/logos/crowdai.png" /></a>
      <a href="https://news.developer.nvidia.com/nvidia-sponsors-learning-to-run-ai-competition-at-nips-2017/"><img src="https://s3.amazonaws.com/osim-rl/logos/nvidia.png" /></a>
      <br />
      <a href="https://aws.amazon.com/blogs/machine-learning/nips-2017-challenge-pushes-deep-learning-to-improve-surgical-outcomes/"><img src="https://s3.amazonaws.com/osim-rl/logos/aws.png" /></a>
      <a href="https://techcrunch.com/2017/08/07/dueling-ais-compete-in-learning-to-walk-secretly-manipulating-images-and-more-at-nips/"><img src="https://s3.amazonaws.com/osim-rl/logos/tc.png" /></a>
      <a href="https://cloud.google.com/"><img src="https://s3.amazonaws.com/osim-rl/logos/gcp.png" /></a>
      <a href="https://www.tri.global/"><img src="https://s3-eu-west-1.amazonaws.com/kidzinski/nips-challenge/tri1.png" /></a>
      <a href="http://opensim.stanford.edu/about/"><img src="https://s3.amazonaws.com/osim-rl/logos/ncsrr.png" /></a>
      
    </div-->
  </div>
</section>
<section class="quickstart">
  <div class="grid">
    <div class="unit one-fifth center-on-mobiles">
      <h3>Get up and running <em>in&nbsp;seconds</em>.</h3>
    </div>
    <div class="unit four-fifths code">
      <p class="title">Quick-start Instructions</p>
      <div class="shell">
        <p class="line">
          <span class="path">~</span>
          <span class="prompt">$</span>
          <span class="command">conda create -n opensim-rl -c kidzik opensim python=3.6.1</span>
        </p>
        <p class="line">
          <span class="path">~</span>
          <span class="prompt">$</span>
          <span class="command">source activate opensim-rl</span>
        </p>
        <p class="line">
          <span class="path">~(opensim-rl)</span>
          <span class="prompt">$</span>
          <span class="command">conda install -c conda-forge lapack git</span>
        </p>
        <p class="line">
          <span class="path">~(opensim-rl)</span>
          <span class="prompt">$</span>
          <span class="command">pip install git+https://github.com/stanfordnmbl/osim-rl.git</span>
        </p>
        <p class="line">
          <span class="path">~(opensim-rl)</span>
          <span class="prompt">$</span>
          <span class="command">python</span>
        </p>
        <p class="line">
          <span class="output">from osim.env import ProstheticsEnv<br/>
env = ProstheticsEnv(visualize=True)<br/>
observation = env.reset()<br/>
for i in range(200):<br/>
&nbsp;&nbsp;&nbsp;&nbsp;o, r, d, i = env.step(env.action_space.sample())</span>
        </p>
      </div>
    </div>
    <div class="clear"></div>
  </div>
</section>
<section class="free-hosting" style="padding-bottom: 1em; padding-top: 1em">
  <div class="grid">
    <h2>NIPS 2017: Learning to Run challenge</h2>
    <p>In 2017 we used osim-rl in a challenge at NIPS were participants were asked to build controllers for running. They did great :)</p>
    <div class="pane-content">

      <div style="width: 100%; padding-bottom: 75%; position: relative;">
	<iframe style="position: absolute; top: 0; left: 0; width: 100%; height: 100%;" src="https://www.youtube.com/embed/rhNxt0VccsE" frameborder="0" allow="autoplay; encrypted-media" allowfullscreen></iframe>
      </div>
    </div>
  </div>
  <div class="clear"></div>

  </div>
</section>