Skip to content

Commit 8432c6c

Browse files
Merge branch 'master' into new-website
2 parents 2888885 + ba1063f commit 8432c6c

File tree

9 files changed

+105
-69
lines changed

9 files changed

+105
-69
lines changed
612 KB
Loading
2.09 MB
Loading

assets/js/footer.js

+2-1
Original file line numberDiff line numberDiff line change
@@ -2,6 +2,7 @@ document.write('<footer class="background-1 color-8 py-3 fs--1 fw-600 font-headi
22
<div class="container">\
33
<div class="row align-items-center">\
44
<div class="col-auto font-1">&copy; 2017-20 Flow Team </div>\
5+
56
<div class="col fs-0 text-right">\
67
<a class="p-2 color-8" target="_blank" href="https://twitter.com/BerkeleyMsl">\
78
<i class="fa fa-twitter"></i>\
@@ -15,4 +16,4 @@ document.write('<footer class="background-1 color-8 py-3 fs--1 fw-600 font-headi
1516
</div>\
1617
</div>\
1718
</div>\
18-
</footer>');
19+
</footer>');

gallery/bottleneck.html

+35-20
Original file line numberDiff line numberDiff line change
@@ -52,16 +52,20 @@ <h1 class="fs-3 fs-md-4">Bottleneck Control</h1>
5252
<div class="fa fa-clock-o fs-1 mr-2 color-3"></div>Dec 2018</div>
5353
<div class="col-4">
5454
<div class="fa fa-file-o fs-1 mr-2 color-3"></div>
55-
<a href="../papers/08569485.pdf">Relevant Paper</a>
55+
<a href="../papers/LagrangianBottlenecks.pdf">Relevant Paper</a>
5656
</div>
5757

5858
</div>
5959
<img class="mb-4 radius-primary" src="../assets/images/experiments/bridge.jpeg">
60-
<p class="lead color-5">Inspired by the rapid decrease in lanes on the San Francisco-Oakland Bay Bridge, we study a bottleneck that merges from four lanes down to two to one.</p>
61-
<p>We demonstrate that the AVs are able to learn a strategy that increases the effective outflow at high inflows, and performs competitively with ramp metering.</p>
62-
60+
<p class="lead color-5">Inspired by the rapid decrease in lanes on the San Francisco-Oakland Bay Bridge, we study a bottleneck, depicted in Fig. 1 that merges from four lanes down to two to one. We come up with a scheme that could save on the costs of deploying traffic light infrastructure by relying on freely available autonomous vehicles.</p>
61+
<figure>
62+
<img src="../assets/images/experiments/bottleneck_control.png" class="img-fluid result" alt="Bottleneck control design">
63+
<figcaption class="small text-center"> Fig. 1, a bottleneck in which four lanes go to two then to 1. Red vehicles are autonomous, white are human. </figcaption>
64+
</figure>
65+
66+
<p>In this system, too large of an inflow into the system leads to irreversible congestion. We demonstrate that we can increase the total flow rate of this system by using AVs. The AVs are able to learn a strategy similar to a traffic light, deciding which lanes go and whch stop. This increases the effective outflow at high inflows, and performs competitively with ramp metering, the solution that is currently deployed on the Bay Bridge. </p>
67+
<p>In the video below, we depict the effects of this congestion.</p>
6368
<div style="margin-bottom: 40px;">
64-
<h3>Videos</h3>
6569
<!--
6670
<a href="https://www.youtube.com/watch?v=Z6QltFAEDeQ?rel=0" class="video-modal btn btn-icon btn-outline-primary btn-icon-left btn-capsule">
6771
<span class="fa fa-youtube">
@@ -75,31 +79,42 @@ <h3>Videos</h3>
7579
<span class="fa fa-youtube">
7680
</span> Play Video 3
7781
</a>
78-
-->
79-
<a href="https://www.youtube.com/watch?v=taqJyMKzPkc&feature=youtu.be" class="video-modal btn btn-icon btn-outline-primary btn-icon-left btn-capsule">
80-
<span class="fa fa-youtube">
81-
</span> Aimsun Video 1
82-
</a>
82+
-->
83+
<iframe width="560" height="315" src="https://www.youtube.com/embed/taqJyMKzPkc" frameborder="0" allow="accelerometer; autoplay; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe>
8384
</div>
84-
85-
86-
87-
88-
<figure>
89-
<img src="../assets/images/experiments/bottleneck_control.png" class="img-fluid result" alt="Bottleneck control design">
90-
<figcaption class="small text-center"> Control structure of the bottleneck. Scale of segments are distorted for visualization. </figcaption>
91-
</figure>
85+
<p>The effects of congestion can be visualized in the following graphic, where we plot the number of entering vehicles per hour vs. the number of exiting vehicles per hour. As can be seen, the exiting number gradually increases but then sharply dips down.</p>
9286

9387
<figure>
9488
<img src="../assets/images/experiments/bottleneck_congestion.png" class="img-fluid result" alt="Bottleneck control design">
95-
<figcaption class="small text-center"> Without control, congestion rapidly forms in the bottleneck.</figcaption>
89+
<figcaption class="small text-center"> Entering vehicles/hour vs. exiting vehicles per hour. Around 1600 entering vehicles per hour, the exiting number begins to decrease.</figcaption>
9690
</figure>
9791

92+
<p>Now we convert 10% of the vehicles to AVs and let them choose their accelerations and actions. As you can see in the attached video, </p>
93+
<div style="margin-bottom: 40px;"> <!--
94+
<a href="https://www.youtube.com/watch?v=Z6QltFAEDeQ?rel=0" class="video-modal btn btn-icon btn-outline-primary btn-icon-left btn-capsule">
95+
<span class="fa fa-youtube">
96+
</span> Play Video 1
97+
</a>
98+
<a href="https://www.youtube.com/watch?v=SoA_7fPJEG8?rel=0" class="video-modal btn btn-icon btn-outline-primary btn-icon-left btn-capsule">
99+
<span class="fa fa-youtube">
100+
</span> Play Video 2
101+
</a>
102+
<a href="https://www.youtube.com/watch?v=5R6GWarVh2o?rel=0" class="video-modal btn btn-icon btn-outline-primary btn-icon-left btn-capsule">
103+
<span class="fa fa-youtube">
104+
</span> Play Video 3
105+
</a>
106+
-->
107+
<iframe width="560" height="315" src="https://www.youtube.com/embed/DQudwzFnW1Y" frameborder="0" allow="accelerometer; autoplay; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe>
108+
</div>
109+
110+
<p>Now we compare the entering vehicles/exiting vehicles curves with and without autonomous vehicles. The curves with 10% autonomous vehicles is depicted in yellow and the 0% autonomous vehicles i.e. the fully human case, is depicted in blue. At high inflows, we see a 25% improvement in the outflow. </p>
111+
98112
<figure>
99113
<img src="../assets/images/experiments/uncontrolVRLBigAxes.png" class="img-fluid result" alt="Bottleneck control design">
100-
<figcaption class="small text-center"> Control structure of the bottleneck; at high inflows the outflow is improved by 25%. </figcaption>
114+
<figcaption class="small text-center">Comparison of the inflow/outflow curves with and without control. </figcaption>
101115
</figure>
102116

117+
<p>For purposes of curiosity, we compare this with a ramp metering strategy in which we add a set of traffic lights at the entrance to the bottleneck. This is the currently deployed control strategy on the Bay Bridge. We see that the two perform comparably at high numbers of entering vehicles, but the traffic lights outperform at low inflows. However, in this low-inflow regime we would not actually apply control as this is before congestion sets in. Thus, we claim that the two schemes perform comparably. </p>
103118
<figure>
104119
<img src="../assets/images/experiments/RLVAlineaBigAxes.png" class="img-fluid result" alt="Bottleneck control design">
105120
<figcaption class="small text-center"> Comparison of inflow, outflow curves for AV control vs. ramp metering. At high inflows they perform comparably. </figcaption>

gallery/index.html

+2-2
Original file line numberDiff line numberDiff line change
@@ -127,10 +127,10 @@ <h5 class="card-title">Bottleneck Control</h5>
127127
</div>
128128
<div class="card-footer">
129129
<div class="media align-items-center">
130-
<a target="_blank" href="../papers/08569485.pdf"><span class="fa fa-file-o"></span> Relevant Paper</a>
130+
<a target="_blank" href="../papers/LagrangianBottlenecks.pdf"><span class="fa fa-file-o"></span> Relevant Paper</a>
131131
<div class="media-body l-h-0">
132132
<a class="color-1" href="../team.html">
133-
<h6 class="mb-0">Posted by Aboudy Kreidieh</h6>
133+
<h6 class="mb-0">Posted by Eugene Vinitsky</h6>
134134
</a>
135135
<span class="fs--1 color-2">Dec 2018</span>
136136
</div>

papers/iccps_2019.pdf

5.08 MB
Binary file not shown.

publications.html

+8-2
Original file line numberDiff line numberDiff line change
@@ -53,7 +53,13 @@ <h1 class="fs-3 mb-3 fw-400"><span class="icon-File-Chart color-primary"></span>
5353
</tr>
5454
</thead>
5555
<tbody>
56-
<tr>
56+
<tr>
57+
<td><b>"Simulation to Scaled City: Zero-Shot Policy Transfer for Traffic Control via Autonomous Vehicles"</b>, K. Jang, E. Vinitsky, B. Chalaki, B. Remer, L. Beaver, A. Malikopoulous, A. Bayen,<em class="color-primary">ICCPS</em>, 2019</td>
58+
<td>
59+
<a href="papers/iccps_2019.pdf" class="btn btn-primary" target="_blank">Download</a>
60+
</td>
61+
</tr>
62+
<tr>
5763
<td><b>"Dissipating stop-and-go waves in closed and open networks via deep reinforcement learning"</b>, A. Kreidieh, C. Wu, A. Bayen, <em class="color-primary">ITSC</em>, 2018</td>
5864
<td>
5965
<a href="papers/08569485.pdf" class="btn btn-primary" target="_blank">Download</a>
@@ -68,7 +74,7 @@ <h1 class="fs-3 mb-3 fw-400"><span class="icon-File-Chart color-primary"></span>
6874
</tr>
6975

7076
<tr>
71-
<td><b>"Benchmarks for reinforcement learning in mixed-autonomy traffic"</b>, E. Vinitsky, A. Kreidieh, L. Flem, N. Kheterpal, K. Jang, C. Wu, F. Wu, R. Liaw, E. Liang, A. Bayen, <em class="color-primary">PMLR, Volume 87</em>, 2018</td>
77+
<td><b>"Benchmarks for reinforcement learning in mixed-autonomy traffic"</b>, E. Vinitsky, A. Kreidieh, L. Flem, N. Kheterpal, K. Jang, C. Wu, F. Wu, R. Liaw, E. Liang, A. Bayen, <em class="color-primary">PMLR, Volume 87, 2018 <span class="badge badge-success">When citing the benchmarks, please cite this paper</span> </em></td>
7278
<td>
7379
<a href="papers/vinitsky18a.pdf" class="btn btn-primary" target="_blank">Download</a>
7480
</td>

team.html

+17-3
Original file line numberDiff line numberDiff line change
@@ -71,6 +71,13 @@ <h5 class="fs-0 mt-3 mb-1">
7171
<span class="color-1">Alben Bagabaldo</span>
7272
</h5>
7373
<div class="fs--1 text-uppercase color-6">PhD student, UC Berkeley</div>
74+
</div>
75+
<div class="col-sm-6 col-md-4 col-lg-3 p-4">
76+
<img class="radius-round px-5" src="assets/images/team/headshot-gilbert.png">
77+
<h5 class="fs-0 mt-3 mb-1">
78+
<a class="color-1" href="https://linkedin.com/in/gilbert-bahati/" target="_blank" >Gilbert Bahati</a>
79+
</h5>
80+
<div class="fs--1 text-uppercase color-6">Undergrad student, UC Berkeley</div>
7481
</div>
7582
<div class="col-sm-6 col-md-4 col-lg-3 p-4">
7683
<img class="radius-round px-5" src="assets/images/team/Joy.jpg">
@@ -120,13 +127,20 @@ <h5 class="fs-0 mt-3 mb-1">
120127
<a class="color-1" href="https://linkedin.com/in/maheshmurag" target="_blank" >Mahesh Murag</a>
121128
</h5>
122129
<div class="fs--1 text-uppercase color-6">Undergrad student, UC Berkeley</div>
130+
</div>
131+
<div class="col-sm-6 col-md-4 col-lg-3 p-4">
132+
<img class="radius-round px-5" src="assets/images/team/headshot-samyak.jpg">
133+
<h5 class="fs-0 mt-3 mb-1">
134+
<a class="color-1" href="https://linkedin.com/in/samyakparajuli" target="_blank" >Samyak Parajuli</a>
135+
</h5>
136+
<div class="fs--1 text-uppercase color-6">Undergrad student, UC Berkeley</div>
123137
</div>
124138
<div class="col-sm-6 col-md-4 col-lg-3 p-4">
125139
<img class="radius-round px-5" src="assets/images/team/headshot-parvate.jpg">
126140
<h5 class="fs-0 mt-3 mb-1">
127141
<span class="color-1">Kanaad Parvate</span>
128142
</h5>
129-
<div class="fs--1 text-uppercase color-6">Undergrad student, UC Berkeley</div>
143+
<div class="fs--1 text-uppercase color-6">Master's student, UC Berkeley</div>
130144
</div>
131145

132146
<div class="col-sm-6 col-md-4 col-lg-3 p-4">
@@ -142,7 +156,7 @@ <h5 class="fs-0 mt-3 mb-1">
142156
<h5 class="fs-0 mt-3 mb-1">
143157
<a class="color-1" href="http://wucathy.com" target="_blank" >Cathy Wu</a>
144158
</h5>
145-
<div class="fs--1 text-uppercase color-6">Assisstant Professor, MIT</div>
159+
<div class="fs--1 text-uppercase color-6">Assistant Professor, MIT</div>
146160
</div>
147161

148162
<div class="col-sm-6 col-md-4 col-lg-3 p-4">
@@ -305,4 +319,4 @@ <h3 class="mb-3">We're always looking for talent</h3>
305319
<script src="assets/js/core.js"></script>
306320
<script src="assets/js/main.js"></script>
307321
</body>
308-
</html>
322+
</html>

0 commit comments

Comments
 (0)