Flow: A Modular Learning Framework for Mixed Autonomy Traffic
Open accessThe DOI matches its Crossref record; 18 of the 18 references it cites were found in a registry or an index.
Abstract
The rapid development ofautonomous vehicles(AVs) holds vast potential for transportation systems through improved safety, efficiency, and access to mobility. However, the progression of these impacts, as AVs are adopted, is not well understood. Numerous technical challenges arise from the goal of analyzing the partial adoption of autonomy: partial control and observation, multivehicle interactions, and the sheer variety of scenarios represented by real-world networks. To shed light into near-term AV impacts, this article studies the suitability of deepreinforcement learning(RL) for overcoming these challenges in a low AV-adoption regime. A modular learning framework is presented, which leverages deep RL to address complex traffic dynamics. Modules are composed to capture common traffic phenomena (stop-and-go traffic jams, lane changing, intersections). Learned control laws are found to improve upon human driving performance, in terms of system-level velocity, by up to 57% with only 4–7% adoption of AVs. Furthermore, in single-lane traffic, a small neural network control law with only local observation is found to eliminate stop-and-go traffic—surpassing all known model-based controllers to achieve near-optimal performance—and generalize to out-of-distribution traffic densities.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Human-level control through deep reinforcement learning
Mnih, V., Kavukcuoglu, K., Silver, D., Rusu, A. A., Veness, J., Bellemare, M. G., Graves, A., Riedmiller, M., Fidjeland, A. K., Ostrovski, G., et al. (2015) · Nature · doi:10.1038/nature14236 · Cited by 30,357
Reinforcement Learning: An Introduction
Sutton, R. S., & Barto, A. G. (1998) · IEEE Transactions on Neural Networks · doi:10.1109/tnn.1998.712192 · Cited by 27,225
Playing Atari with Deep Reinforcement Learning
Mnih, V., Kavukcuoglu, K., Silver, D., Graves, A., Antonoglou, I., Wierstra, D., & Riedmiller, M. (2013) · arXiv · doi:10.48550/arxiv.1312.5602 · Cited by 9,471
MuJoCo: A physics engine for model-based control
Todorov, E., Erez, T., & Tassa, Y. (2012) · 2012 IEEE/RSJ International Conference on Intelligent Robots and Systems · doi:10.1109/iros.2012.6386109 · Cited by 4,767
Congested traffic states in empirical observations and microscopic simulations
Treiber, M., Hennecke, A., & Helbing, D. (2000) · Physical Review E · doi:10.1103/physreve.62.1805 · Cited by 4,762 · cited 3×
“The Intelligent Driver Model (IDM) is a car following model capable of accurately representing realistic driver behavior [23] and reproducing traffic waves, and is commonly used in the transportation research community.”
BackgroundMethodTrust region policy optimization
Schulman, J., Levine, S., Moritz, P., Jordan, M. I., & Abbeel, P. (2015) · arXiv · doi:10.48550/arxiv.1502.05477 · Cited by 4,578
End to End Learning for Self-Driving Cars
Bojarski, M., Del Testa, D., Dworakowski, D., Firner, B., Flepp, B., Goyal, P., Jackel, L. D., Monfort, M., Muller, U., Zhang, J., et al. (2016) · arXiv · doi:10.48550/arxiv.1604.07316 · Cited by 3,198
Traffic Flow Prediction With Big Data: A Deep Learning Approach
Lv, Y., Duan, Y., Kang, W., Li, Z., & Wang, F.-Y. (2014) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2014.2345663 · Cited by 3,000
A solution to the simultaneous localization and map building (SLAM) problem
Dissanayake, M. W. M. G., Newman, P., Clark, S., Durrant-Whyte, H. F., & Csorba, M. (2001) · IEEE Transactions on Robotics and Automation · doi:10.1109/70.938381 · Cited by 2,661
Bellman, R. (1957) · Indiana University Mathematics Journal · doi:10.1512/iumj.1957.6.56038 · Cited by 2,408 · cited 1×
“…in this article tackles scenarios which conform to the standard interface of an episodic finite-horizon discounted Markov decision process (MDP) [94], [95], defined by the tuple ( SS, A , P, r, ρ 0 , γ, T ) , where SS is a (possibly infinite) set of states, A is a (possibly infinite) set of…”
BackgroundStanley: The robot that won the DARPA Grand Challenge
Thrun, S., Montemerlo, M., Dahlkamp, H., Stavens, D., Aron, A., Diebel, J., Fong, P., Gale, J., Halpenny, M., Hoffmann, G., et al. (2006) · Journal of Field Robotics · doi:10.1002/rob.20147 · Cited by 2,237 · cited 1×
“Existing tools are largely limited to two commonly studied regimes: where AVs are few enough as to not affect the surrounding traffic dynamics [3], [4], [5], or so ubiquitous as to become a coordination problem [6], [7], [44].”
MethodThe Arcade Learning Environment: An Evaluation Platform for General Agents
Bellemare, M. G., Naddaf, Y., Veness, J., & Bowling, M. (2013) · Journal of Artificial Intelligence Research · doi:10.1613/jair.3912 · Cited by 2,161 · cited 1×
“While RL testbeds have enabled significant progress for algorithm development [99], [100], [101], testbeds which enable RL to shed light into important real-world problem domains remain limited.”
BackgroundBenchmarking Deep Reinforcement Learning for Continuous Control
Duan, Y., Chen, X., Houthooft, R., Schulman, J., & Abbeel, P. (2016) · arXiv · doi:10.48550/arxiv.1604.06778 · Cited by 1,114 · cited 1×
“We use the Trust Region Policy Optimization (TRPO) [82] method, with linear feature baselines as described in Duan, et al. [105], discount factor γ = 0 .”
MethodHelp or hindrance? The travel, energy and carbon impacts of highly automated vehicles
Wadud, Z., MacKenzie, D., & Leiby, P. (2016) · Transportation Research Part A: Policy and Practice · doi:10.1016/j.tra.2015.12.001 · Cited by 859 · cited 2×
“Autonomous vehicles (AVs) are projected to enter society in the very near future, with full adoption in select areas expected as early as 2050 [1].”
BackgroundGated Feedback Recurrent Neural Networks
Chung, J., Gulcehre, C., Cho, K., & Bengio, Y. (2015) · arXiv · doi:10.48550/arxiv.1502.02367 · Cited by 583 · cited 1×
“The GRU is a recurrent neural network that makes use of parameterized update and reset gates, which enable decision making based on both current and past inputs [98].”
BackgroundBrockman, G., Cheung, V., Pettersson, L., Schneider, J., Schulman, J., Tang, J., & Zaremba, W. (2016) · arXiv · doi:10.48550/arxiv.1606.01540 · Cited by 582 · cited 2×
“While RL testbeds have enabled significant progress for algorithm development [99], [100], [101], testbeds which enable RL to shed light into important real-world problem domains remain limited.”
BackgroundMethodSafe, Multi-Agent, Reinforcement Learning for Autonomous Driving
Shalev-Shwartz, S., Shammah, S., & Shashua, A. (2016) · arXiv · doi:10.48550/arxiv.1610.03295 · Cited by 509 · cited 1×
“The canonical example is the performance of a single AV, compared to a human driver [9], [10].”
BackgroundRLlib: Abstractions for Distributed Reinforcement Learning
Liang, E., Liaw, R., Moritz, P., Nishihara, R., Fox, R., Goldberg, K., Gonzalez, J. E., Jordan, M. I., & Stoica, I. (2017) · arXiv · doi:10.48550/arxiv.1712.09381 · Cited by 473
Metadata
- Authors
- Wu, C., Kreidieh, A. R., Parvate, K., Vinitsky, E., & Bayen, A. M.
- Venue
- IEEE Transactions on Robotics
- Published
- 2022-04
- Volume
- 38
- Issue
- 2
- Pages
- 1270-1286
- ISSN
- 1552-3098 (ISSN-L)
- 1941-0468
- OpenAlex
- W3193464862
- Cited by
- 193
- References
- 18
- FWCI
- 13.28
- Type
- journal article
- Language
- en
- Open access
- green
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.