You are here

State releases automated vehicle considerations for rural environments

COLUMBUS __ In a little under a year, DriveOhio’s Rural Automated Driving Systems (ADS) project collected nearly 60 terabytes of data that will help define future technology needs for automated vehicles in rural environments. While automated driving systems have the potential to dramatically improve roadway safety, data shows that these systems must be refined to maximize that potential in rural areas.

“The data this project gathered is invaluable. We predicted certain challenges before the project began, but to now have data that clearly demonstrates the performance inconsistencies marks the first step in developing the solution,” said DriveOhio Executive Director Preeti Choudhary.

Most automated vehicle testing to date has been conducted in urban areas without the hilly and curvy terrain, limited sight distances, and shaded tree canopies typical of a rural environment.  Funded in part by a $7.5 million grant from the U.S. Department of Transportation, this project accomplished one of the most comprehensive testing efforts yet to be conducted on rural roads in the United States.

EASE Logistics was the host fleet partner for one deployment featuring a pair of semi-trucks equipped with platooning technology that enabled them to operate in a ‘follow the leader’ scenario. While the driver of the lead truck controlled the speed, the following vehicle automatically responded to the lead vehicle’s movement. With no input from the second driver, the technology mimicked the speed of the lead vehicle, with precisely matched acceleration and braking.

In total, the pair of tractor-trailers operated by EASE Logistics made more than 100 deliveries to customers traveling nearly 44,000 miles throughout Ohio, which includes more than 11,000 miles in platooning mode. Though no official fuel data was collected or analyzed, EASE estimates about a 10 percent fuel savings during the deployment. Additionally, the driver of the following truck expressed that the technology reduced stress while driving.

“The implementation of semi-automated trucks holds immense potential for enhancing roadway safety, optimizing transportation efficiency, and significantly reducing emissions across the U.S.” said Peter Coratola, Jr., president and CEO of EASE. “We’re pleased with the results and believe these findings will be critical to delivering innovative and practical solutions that address the real-world challenges faced by supply chain managers today.”

Another deployment showcased three passenger vehicles equipped with technology capable of controlling steering, acceleration, and braking, including the ability to perform lane changes and navigate intersections. Drivers from the Transportation Research Center, Inc. and Ohio University tested the technology on divided highways and rural two-lane routes in Athens and Vinton County. 

“By deploying these vehicles, we were able to collect meaningful on-road data and see first-hand how automated vehicles may operate in rural environments,” said Taylor Manahan, executive director of services at TRC. “Thanks to this project, innovators will have important information to help them advance new technologies, while understanding the challenges and opportunities of rural transportation systems. We were proud to apply the expertise of our engineers, researchers and technical staff to help with deployment of these vehicles to help advance the future of transportation in rural communities.”

Since the passenger vehicles operated a higher level of automation than the truck platoon, the technology relied on high-definition models of the routes, which were created and verified prior to deployment. The vehicles were also equipped with external cameras and sensors for object detection.

During the project, the passenger vehicles made 331 trips along several pre-determined routes, covering more than 5,000 total miles, including nearly 3,000 miles in automated mode. The majority of the trips – about 75 percent – experienced disengagements of the technology, meaning that either the driver or the software disconnected the automated driving system and operated the vehicle manually. The most common disengagements involved a loss of GPS or cellular service, object detection, and the behavior of surrounding traffic or pedestrians.

Since the software required the models to be created in advance, the driver was not able to make adjustments to how the vehicle operated in real-time based on conditions like heavy traffic, weather, or work zones. Human driving behaviors are typically adjusted for these factors, so disengaging the technology and taking over manually was the only way to adapt to surrounding traffic behavior.

“From a safety driver's perspective for autonomous vehicles, it took patience and time to adapt to the vehicle that would slow down on its own for curves and come to a stop at lights. The systems would always operate on the most efficient slow down path, even though a driver might brake earlier,” said Dr. Jay Wilhelm, Associate Professor of Mechanical Engineering at Ohio University. “The disengagements by drivers were sometimes due to us humans applying more safety than needed. As time went on, we learned to trust the machine more and as a result less disengagements occurred and more data was gathered.”

Data collected throughout the deployments includes the vehicle’s position, speed, steering inputs, and acceleration along with information about the driving environment like the speed and position of other vehicles, traffic signal detection, and object identification. Additionally, disengagement data was captured each time the automated driving system turned off.

Students at Youngstown State University then analyzed this data as part of a year-long course. By using artificial intelligence and machine learning to log each disengagement and the features present at each instance, students aided the project’s understanding of the events and their overall significance. While no cause-and-effect conclusions were drawn from the data as part of the project, the team’s effort demonstrates the value of future data analysis.

The final report for the $13.4 million Rural Automated Driving Systems project can be found here: Ohio Rural Automated Driving Systems (ADS) Project Final Evaluation Report

Section: