Abstract
Shockwaves are a naturally emerging phenomena in freeway traffic, but they represent one of the largest safety risks on freeways. Freeway drivers do not expect to encounter abrupt drops in speed or stopped traffic, as a result, shockwaves sharply increase the accident rates, particularly in the context of rear end collisions. For example, US interstate highways in 2021 saw the following rear-end collision numbers: Fatality 985, Injury-Only 71,408, Property-Damage-Only 152,011. Rear end collision severity is directly related to the relative speed between the involved vehicles, shockwaves increase these relative speeds, and thus, they also increase accident severity. Shockwaves also reduce freeway capacity and have a detrimental impact on traffic flow.
Connected and autonomous vehicles (CAV) hold the promise to attenuate and eliminate shockwaves (and thus, also reduce the severity and number of accidents), but only if the system is explicitly designed to do so. The very factors that give rise shockwaves in human driven vehicles (HDV) will also do so in CAV. While CAV offer new ways to manage traffic dynamics, an automated freeway will still be subject to traffic dynamics. The real challenge is designing the CAV system so that it ensures the safest possible operation, and then within those bounds, the greatest operational efficiency (maximizing capacity, minimizing delays, etc.).
This proposal builds from earlier Safety21 research. Phase 1 developed the transition from unstable stop and go traffic to a smoother trajectory for the first CAV. It uses current conditions along the corridor to estimate the optimal trajectory for the CAV encountering congestion. Phase 2 extended the control to a platoon of CAV, suppressing major disturbances, typically with the 15th vehicle exhibiting a nearly constant speed trajectory even while the 1st vehicle in the platoon encounters stop waves.
The proposed research for Phase 3 seeks to address several key aspects. First, extend the methodology to actively manage multi-lane flows, including sequencing and optimizing lane change maneuvers. In all phases of this work, the HDV dynamics ahead of the first CAV come from empirical microscopic data sets. Phase 2 simply accepts the lane changes when and where the HDV happened to change lanes. But these maneuvers disrupt the platoons and degrade performance. Phase 3 will actively handle the lane change maneuvers so that they no longer disrupt platoons and in fact, can be used to robustly support the platoons. Another limitation of Phase 2 is that the empirical data only covers 1/3 of a mile, as a result, it is difficult to match demand of vehicles from upstream with the variable supply of capacity downstream. This research will develop ways to use multiple platoons to buffer against capacity fluctuations and provide finer control, e.g., dynamically advancing vehicles from upstream platoons when downstream capacity exceeds forecasts and slowing platoons when it falls below forecasts. In this way, the proposed research anticipates and develops the management strategies needed to maintain the smoothly flowing queued traffic with CAV.
Description
Timeline
Strategic Description / RD&T
Section left blank until USDOT’s new priorities and RD&T strategic goals are available in Spring 2026.
Deployment Plan
July – September 2025
1. quarterly report
October – December 2025
1. quarterly report
January – March 2026
1. quarterly report
April – June 2026
1. final report
2. conference and/or journal paper submission
Expected Outcomes/Impacts
Since the earliest autonomous vehicle research the focus has primarily been the very challenging task of making the technology work. While the fact that automated vehicles would benefit operations was taken as self-evident. To date, these priorities were well placed, but now that we are at the cusp of finally deploying safe and reliable CAV, it is important to take a step back and critically assess exactly how CAV can and should impact operations. Much of the past motivation was built on the recognition that CAV can maintain tighter headways, which increases vehicle throughput locally. This research seeks to take a step back from such a point perspective, moving to a proactive corridor perspective. Freeway operations are ultimately constrained by a few links that either have the lowest capacity or highest demand. Away from these bottlenecks is excess local capacity that we plan to use to optimize the flow of traffic through the bottlenecks. This research seeks work outward from bottlenecks to both supply demand to meet (potentially dynamic) capacity and manage any excess demand in stable manner where the waiting vehicles do not give rise to shockwaves. By mitigating shockwaves it will improve safety and increase throughput.
Expected Outputs
It is anticipated that the quantifiable outputs of this research will include one Master's thesis and several scholarly publications. While harder to quantify, we believe the goal oriented approach, "what do we want CAV to do and how do we get there?" will help the work achieve impactful results. As discussed in the body, we seek to move from simply addressing disturbances as they arise to proscribing the target traffic state over time and space to prevent the disturbances from arising, which will address the primary objective of improving safety (eliminating unexpected speed drops), while also leading to secondary benefits of increasing throughput (stable traffic has a higher capacity than fluctuating traffic states).
TRID
A TRID search for related research projects turned up many related projects, though none that directly overlap the proposed research. The following reviews the closest projects, full reference for the numbered citations can be found in the included document.
There are many broad scale projects in which this topic might fit, but none specifically call out this topic, e.g., [1-5]. There are several projects looking into the general human interface or interaction with CAV that will help set the context, e.g., [6-7]. There are studies that are looking into modeling the car following behavior of CAV with an emphasis on the autonomous vehicle aspects [8], others emphasize the communication aspect for vehicle routing [9].
Several projects hold promise for end-to-end compatibility with the proposed: developing simulation platforms [10], modeling microscopic vehicle interactions for maneuvering without specific consideration of the macroscopic dynamics [11-13]. The closest related studies to our proposal consider using CAV to harmonize speed or dampen disturbances [14-16], with the key difference that those studies seek to address disturbances as they arise while we seek to proscribe the target traffic state over time and space to prevent the disturbances from arising.
Individuals Involved
| Email |
Name |
Affiliation |
Role |
Position |
| coifman.1@osu.edu |
Coifman, Benjamin |
OSU |
PI |
Faculty - Tenured |
| redmill.1@osu.edu |
Redmill, Keith |
The Ohio State University |
Other |
Other |
Budget
Amount of UTC Funds Awarded
$73762.00
Total Project Budget (from all funding sources)
$137891.00
Documents
Match Sources
No match sources!
Partners
| Name |
Type |
| DriveOhio |
Deployment Partner_ Deployment Partner_ |
