Skip to main content
eScholarship
Open Access Publications from the University of California

Scholarly Works (3 results)

  • Thesis
  • Peer Reviewed
UC Berkeley Electronic Theses and Dissertations (2024)

With the advent of autonomous vehicles (AVs) comes a broad array of possibilities for control. Looking beyond the immediate wave of research that is focused on training models that can drive safely, this work looks into the future, and aims to develop models for AVs that can achieve more than safe driving. Traffic dynamics are notoriously difficult to model and capture on the micro-level, with behaviors ranging from human-observable to ones we are not aware of, happening every second. Reinforcement learning (RL) is a method which is effective in capturing structure from highly complex, heavy, behavioral data. In this work, we use RL and leverage its ability to understand complex human-driver and traffic dynamics in order to develop policies that are able to not only drive, but drive in a way that can smooth traffic.

With the goal of taking these traffic-smoothing algorithms to the real world, the aim of this work takes a path through three parts, from work done purely in simulation to a eventual 100-AV road test. We first explore the concept of using RL as a means of control for wave-smoothing policy control by examining experiments across a variety of traffic scenarios that demonstrate its effectiveness. This portion happens purely in simulation and explores various components of RL design, from environment design to reward shaping. With the goal of deployment always in mind, we also conduct research on how to develop RL policies that are robust enough to survive the transfer from simulation to the real world while sacrificing minimal performance. Lastly, the work comes together to explore the development and deployment of the MegaVanderTest, the deployment of 100 RL-enabled AVs, and to our knowledge, the largest test of AVs designed to smooth traffic.

    • Article
    Policy Briefs (2022)
    In an ideal world, all cars along a congested roadway would travel at the same constant average speed; however, this is hardly the case. As soon as one driver brakes, trailing cars must also brake to compensate, leading to “stop and go” traffic waves. This unnecessary braking and accelerating increases fuel consumption (and greenhouse gas emissions) by as much as 67 percent.1 Fortunately, automated vehicles (AVs) — even Level 2 AVs2 which are commercially available today — have the potential to mitigate this problem. By accelerating less than a human would, an AV with flow smoothing technology is able to smooth out a traffic wave, eventually leading to free-flowing traffic (See Figure 1). To demonstrate the potential of flow smoothing on reducing greenhouse gas emissions, researchers at UC Berkeley used a calibrated model of the I-210 freeway in Los Angeles to simulate and measure the effect of deploying different percentages (10%, 20%, 30%) of flow-smoothing AVs on the average miles per gallon (MPG) of non-AVs in the traffic system.
      Cover page: Automated Vehicle Technology Has the Potential to Smooth Traffic Flow and Reduce Greenhouse Gas Emissions
      • Article
      ITS reports (2022)
      Passenger and heavy-duty vehicles make up 36% of California’s greenhouse gas (GHG) emissions. Reducing emissions from vehicular travel is therefore paramount for any path towards carbon neutrality. Efforts to reduce GHGs by encouraging mode shift or increasing vehicle efficiency are, and will continue to be, a critical part of decarbonizing the transportation sector. Emerging technologies are creating an opportunity to reduce GHGs. Human driving behaviors in congested traffic have been shown to create stop-and-go waves. When waves form, cars periodically slow down (sometimes to a stop) and then speed back up again; this repeated braking and accelerating leads to higher fuel consumption, and correspondingly increasingly GHG emissions. Flow smoothing, or the use of a specially designed adaptive cruise controllers to dissipate these waves, can reduce fuel consumption of all the cars on the road. By keeping all vehicles at a constant speed, flow smoothing can minimize system-wide GHG emissions. This report presents the results of flow-smoothing when used in simulation, discusses current work on implementing flow-smoothing in real world-highways, and presents policy discussions on how to support flow smoothing.
        Cover page: Using Automated Vehicle (AV) Technology to Smooth Traffic Flow and Reduce Greenhouse Gas Emissions
        Top