Breakthroughs to The Future

by David Kuehn and Mark Fitzgerald

Emerging technologies promise transformational changes in transportation’s efficiency, safety, and environmental impacts. Leading the charge? FHWA’s EAR Program.

FHWA’s Exploratory Advanced Research (EAR) Program is conducting big-picture research with the potential to produce transformational improvements on the Nation’s highways, especially metropolitan interstates such as this one in northern Virginia.

Big problems demand big solutions. Few would dispute that today’s highways have big problems: clogged roads, too many fatalities and injuries, and too much air pollution caused by stop-and-go traffic.

Certainly, a number of targeted solutions like lanes for high-occupancy vehicles or pedestrian crossing islands are chipping away at these problems, especially in terms of improving safety. Through applied research, the Federal Highway Administration (FHWA) and State departments of transportation (DOTs) have developed specific approaches to mitigating congestion and the resulting pollution, as well as enhancing safety. Implementing those individual measures has resulted in immediate benefits in all three areas.

To solve the seemingly intransigent, large-scale problems, however, requires big-picture thinking, out-of-the-box approaches, and long-term research. Studies are needed to develop emerging technologies that offer solutions outside normal transportation channels.

To conduct this high-risk, long-horizon research, and to bridge the gap between basic and applied research, Congress authorized FHWA’s Exploratory Advanced Research (EAR) Program in 2005 under the Safe, Accountable, Flexible, Efficient Transportation Equity Act: A Legacy for Users. Congress reauthorized the EAR Program in 2012 under the Moving Ahead for Progress in the 21^st Century Act.

The EAR Program addresses the need to conduct breakthrough research with the potential for transformational improvements. Findings from projects funded by the EAR Program could range from new research methods and models to new system concepts and prototypes. The program is generating innovative insights that could change the way the transportation community looks at the big problems.

One of the program’s focus areas--connected highway-vehicle systems--holds the promise of becoming the foundation for the future of roadway transportation. To realize that ambitious goal, connected-vehicle systems depend on a set of emerging technologies. Projects funded by the EAR Program in this focus area have leveraged breakthroughs in sensor and control technology, wireless communications, data fusion, and information sciences.

In its first 8 years, the EAR Program initiated 6 solicitations and awarded 52 projects (37 of which were ongoing as of August 2013) involving government, academic, and private sector researchers. These projects represented the investment of $43 million in FHWA funds and leveraged $17 million in matching State, local, academic, private sector, and nongovernmental funds. The program will fund additional projects throughout 2014.

Three EAR-funded projects--freeway merge assistance, selected mobility applications, and advanced traffic signal control algorithms--are discussed in this article. The freeway merge assistance project involved developing a lane-level variable speed limit algorithm, a lane-changing advisory algorithm, three gap-responsive onramp signal algorithms, a merge-control algorithm, and steps for system integration and evaluation. The project on selected mobility applications focused on cooperative adaptive cruise control, variable speed limits, automated truck platoons, and cooperative vehicle-highway automation systems. The project on advanced traffic signal control algorithms examined how vehicles and infrastructure could leverage real-time information to improve mobility and safety and reduce emissions.

Through partnerships with the private sector, academia, and government laboratories, the EAR Program has produced findings that are advancing these emerging technologies. The goal is to encourage further research and development that can lead to implementation by the highway industry.

Freeway Merge Assistance

To address bottlenecks and safety concerns at freeway merge areas, the EAR Program invested in research that identified three critical steps needed to implement a system of assistance at these interchanges. The first step involves researching how to create discrete gaps in the rightmost lanes of freeway mainlines (the high-speed segments of highways), as they approach onramps. The second and third steps involve communicating the availability of gaps to drivers waiting to merge onto the freeway mainlines and commanding the merging action through real-time communications and control actions.

In 2009, the EAR Program awarded the University of Virginia Center for Transportation Studies a cooperative agreement to develop these steps. Named the Freeway Merge Assistance project, the research involved enhancing the results from several previous projects funded by FHWA, the National Science Foundation, the Virginia DOT, and the California DOT (Caltrans).

Focusing on the three critical steps, the researchers established a simulation of connected-vehicle traffic and communications. The simulation replicated precise vehicular movements and incorporated wireless communications based on standards for Wireless Access in Vehicular Environments (WAVE) and dedicated short range communications (DSRC) in a connected-vehicle environment.

The research team also simulated real connected-vehicle message sets, which support interoperability among safety, mobility, and sustainability applications. The message sets can be sent for wireless access in vehicles. Examples might be messages about emergency vehicles, travel advisories, and weather and road conditions.

The research team found that a merge assistance system has the potential to significantly improve freeway operations in a connected-vehicle environment. The technology enables safe merging in small spaces at the highest speed possible. With connected-vehicle communications in effect, the merging control algorithm is especially promising, generating significant improvements.

Findings from this project (detailed below) are expected to help transportation agencies leverage the many benefits of connected-vehicle technology. Some benefits include improved merging, reductions of emissions, and enhanced performance of the transportation network.

The Virginia researchers also developed algorithms for lane-level variable speed limits, lane-changing advisories, three gap-responsive onramp signals, and merging control, plus steps for system integration and evaluation. Each is described below, briefly.

Lane-Level Speed Control

The lane-level variable speed limit created by the Virginia team is an algorithm that implements a lower speed limit on the rightmost lane of a freeway mainline. The algorithm determines a new speed limit based on the densities in the left and rightmost lanes to encourage mainline vehicles to move to the left, creating a better situation for drivers who need to merge from onramps. The variable speed limits are communicated through invehicle display messages.

Evaluations showed that the algorithm can improve network performance slightly, with up to 7 percent higher travel speeds. An individual lane-level analysis revealed an increase of 9 percent in average speed in the rightmost lanes of freeway mainlines within merge areas.

“We found that the algorithm has the potential to create better merging situations and improved freeway operations,” says Hyungjun Park, a senior research scientist at the University of Virginia Center for Transportation Studies, who worked on the project. “We expect that these improvements might enhance driver safety by reducing differences in speeds and density within a merge area.”

Lane-Changing Advisory

The lane-changing advisory algorithm developed by the Virginia researchers calculates the anticipated sizes of the gaps for the lead and lag (following) cars and trucks, using equations based on accelerating, maintaining current speeds, and decelerating. To create more space within the ramp merging area, the algorithm provides lane-changing advisories to freeway mainline vehicles based on the calculated variable gap sizes.

“The algorithm advises drivers about the right time to make a lane change,” explains Park. “If there is a gap available in the left lane for a vehicle that is traveling in the rightmost lane, an invehicle communication device displays a message advising the driver to merge left. This allows for more space in the rightmost lane.” It also improves safety by decreasing lane-changing conflicts.

The researchers found that the algorithm has the potential to address freeway merge conflicts. They observed that the average speed of vehicles in freeway mainlines increased by 6 percent within merge areas, an encouraging operational improvement. Algorithm performance was sensitive to driver compliance rates, however, suggesting that a rate of 90 percent or higher is necessary for the proposed algorithm to achieve these results.

Gap-Responsive Onramp Signal

Three algorithms for gap-responsive onramp signals estimate the target “time step,” or the time when a vehicle on the ramp reaches the merging point after passing the ramp’s traffic signal. The algorithms then determine gap availability in the target time step by predicting locations of mainline vehicles. Lastly, the algorithms trigger a red or green signal on the onramp based on gap availability.

“The algorithms determine the expected gap availability within the merging area of the mainline freeway,” explains Park. “They control the start time of the onramp vehicle. If there is a gap available, they provide a green light that tells the onramp vehicle to go ahead and merge. If there isn’t a gap available, a red light tells the vehicle to stop and wait.”

Researchers had difficulty estimating target time steps and predicting mainline vehicle locations, but they achieved significant environmental benefits: a reduction of 8 percent in fuel consumption and 16 percent in carbon monoxide emissions.

Merging Control

The aim of the freeway merging control algorithm is cooperatively (with the drivers) to control leading mainline vehicles, lagging mainline vehicles, and ramp merging vehicles. By using detailed data and control or driver advisory capabilities enabled by connected-vehicle technology, the researchers found that smoother merging is possible.

“If you’re the leading vehicle on the mainline, you may need to accelerate a little bit, and the lagging vehicle on the mainline may need to decelerate a little bit,” Park adds. “If you’re the onramp vehicle, you may have to maintain your speed, accelerate, or decelerate a little bit. Out of all of these possibilities, the most desirable actions for all vehicles involved are determined based on the detailed vehicle data from the connected-vehicle system.” Drivers would receive merging guidance on invehicle display screens.

Under certain levels of congestion, rates of compliance, and connected-vehicle market penetration, evaluations demonstrated that the control algorithm has significant potential to improve freeway merging operations, thus improving traffic efficiency and safety. Computer simulation tests showed that the control strategy can improve average network speed by 42 percent, reduce total travel time by 26 percent, and decrease total delay time by 55 percent. The algorithm also reduced crossing conflicts by 8 percent, rear-end conflicts by 70 percent, and the total number of conflicts by 45 percent.

System Integration And Evaluation

The Virginia researchers evaluated each algorithm using results from a simulation of connected-vehicle communications. They then integrated the algorithms with the evaluations.

They found that the algorithm for lane-level variable speed limits was not effective when considering real conditions for communications of connected-vehicle message sets. Also, taking into account connected-vehicle communications, the lane-changing advisory algorithm was unable to improve mobility, due to occasional unreliability in communications.

In the integrated simulation of connected-vehicle traffic, however, the merge control algorithm produced benefits. They included an increase of 24 percent in average speeds, a decrease of 12 percent in total travel time, and a decrease of 18 percent in total delay time.

“In the first place, we assumed that in a connected-vehicle environment, all of the messages would be successfully transmitted, which may not be the case in reality,” says Park. “The bottom line is that we can achieve significant benefits from these algorithms.” Park is referring to simulation results and calculations that assume real-time communications will work as intended. He concludes, “Yet, at the same time, we need to consider real connected-vehicle communications to be able to precisely estimate these benefits.”

Selected Mobility Applications

Congestion and limited highway capacity have spurred FHWA to pursue solutions through intelligent vehicle and highway infrastructure applications. In 2007, the EAR Program funded a project focused on developing and evaluating selected mobility applications for connected-vehicle technology (previously known as vehicle-infrastructure integration technology).

In cooperation with Caltrans, researchers from the Partners for Advanced Transportation TecHnology (PATH) program at the University of California, Berkeley, developed new strategies to reduce bottlenecks, minimize congestion, and maximize throughput by using the capabilities of intelligent vehicles and highway infrastructure.

Completed in July 2011, the vehicle-infrastructure integration study found that new connected-vehicle and highway systems can lead to substantial safety, operational, and environmental benefits. The project team increased the size and effectiveness of lane capacities using cooperative adaptive cruise control (CACC), which employs vehicle control technology through advanced communications. The PATH researchers modeled, tested, and demonstrated prototype systems to improve traffic flow by calculating variable speed limits and communicating them to drivers. They also demonstrated the feasibility of reducing fuel consumption and raising the capacity of truck-only lanes by using automated truck platoons.

Cooperative Adaptive Cruise Control

The PATH researchers used wireless communication technology to test CACC applications that have the potential to increase lane capacity while improving driver trust of automated technology--willingness to let the automobile determine an appropriate travel speed and safe distance behind the vehicle ahead. By using vehicle-to-vehicle communications, the researchers found that CACC offers greater benefits than commercially available adaptive cruise control.

The adaptive cruise control that is commercially available has radar that scans ahead and reduces the following vehicle’s set speed to allow it to follow a slower moving vehicle more closely. This enables a safe following gap between vehicles but is not as effective as CACC at increasing lane capacities, according to Steven Shladover, the PATH program research engineer who led the project.

Employing vehicle-to-vehicle communications, a vehicle with CACC can receive data from any of the vehicles in front of it, not just the car immediately in front of it. This means vehicles can anticipate speed changes, making traffic dynamics more stable, so disturbances are not amplified from one car to the next. The CACC system thus provides more opportunities to avoid crashes.

“With CACC, we were able to get a more accurate and tighter vehicle following behavior,” explains Shladover. “This stable vehicle following behavior is a big improvement over the commercially available adaptive cruise control, which is not stable when you put a string of cars together that are equipped with that technology.”

This touch-screen control panel enabled researchers to adjust the speed and following distance of a vehicle using cooperative adaptive cruise control. The touch screen was used for tests only, not for actual deployment.

He adds, “It’s possible to get vehicles closer together with CACC. We found that the gaps between vehicles that drivers were comfortable selecting could be less than half the size of the gaps with the commercially available system.

“With the cooperative system, as soon as you see the brake lights on the car ahead, you feel your car braking immediately. Because you feel this response, you gain confidence that your car knows what’s going on with the car in front of you. It feels safe to be operating this close because the car is responding so quickly to what’s happening ahead--and, of course, it is safe,” Shladover adds.

The researchers incorporated data provided by drivers in field experiments into a traffic simulation. They then demonstrated that if all the vehicles on the road were equipped with CACC technology and drivers chose the gap distribution used in the experiment, the freeway lane capacity would nearly double.

“It’s a big difference,” observes Shladover. “It’s just short of 4,000 vehicles per hour per lane in a simulation assuming 100 percent market penetration.”

Variable Speed Limits

Also under the CACC project, PATH researchers examined mobility applications that involved modeling, testing, and demonstrating a prototype system to improve traffic flow by adjusting speed limits and communicating them to drivers.

“We did a lot of work in a computer simulation,” says Shladover. “We wanted to understand how reduced speed limits would provide higher capacity through a bottleneck area on a freeway where you have some traffic congestion because of a lane drop or a weaving section.”

The PATH team also conducted real-world tests on a section of I–80 in California where a bottleneck occurs because drivers entering from a right onramp have to make multiple lane changes in time to reach a left exit.

“If the upstream traffic is slowed before it gets there, then it is easier to do those lane changes and have the traffic streams weaving together,” says Shladover, “and it’s also safer because of the reduced speed differential.”

Shown here is a platoon of six sedans using cooperative adaptive cruise control. The technology enables a stable vehicle-following behavior that has the potential to increase lane capacity.

The researchers conducted an experiment to determine an appropriate speed limit for enabling easier lane changes on the bottleneck section. They then communicated this speed to a test vehicle moving through that section. The team found that variable speed limits offered significant potential to prevent traffic delays. In simulations on I–80, they demonstrated favorable results in selecting speeds calculated to prevent breakdowns of traffic flow.

FHWA is continuing to analyze the potential of variable speed limits under a study sponsored by its Saxton Transportation Operations Laboratory at the Turner-Fairbank Highway Research Center in McLean, VA. Under this study, the researchers are working on experiments designed to mitigate congestion caused by merge areas and lane drops on I–66 in northern Virginia.

“We’re developing an experiment in which this type of reduced speed limit can be communicated to several test vehicles, which use it as the set speed for their adaptive cruise control systems,” says Shladover.

Automated Truck Platoons

Also under the EAR Program, PATH conducted automated control tests for truck platoons. The goals were to reduce fuel consumption and to increase the capacity of truck lanes. Using a wireless communications system, the researchers successfully coordinated a platoon of three tractor-trailer trucks traveling at 53 miles per hour, mi/h (85 kilometers per hour, km/h) in various maneuvers involving lanes that joined or split.

“We were able to get three trucks as close as [13 feet] 4 meters apart while they were traveling at 53 miles per hour,” says Shladover. “That’s a very short gap. We made a series of fuel consumption measurements while the trucks were driven at various gaps in the platoon and also driven individually.”

Researchers measured fuel savings of 10 to 14 percent for the following trucks and 4 to 5 percent for the leading truck. These savings are especially meaningful to the owners of truck fleets, as fuel comprises a major operating expense.

“We consider this a more advanced version of the cooperative adaptive cruise control because we run the trucks much closer together and at a constant spacing,” Shladover adds. “And if you have a dedicated truck lane, you don’t have to worry about cars cutting into the middle of the platoon. This also means the lead truck can maintain a more constant speed without having to deal with the changing speeds of other vehicles on the road.”

Cooperative Vehicle-Highway Automation Systems

What can the U.S. transportation community learn from cooperative vehicle-highway automation systems that are being developed abroad? Research on these systems has advanced significantly in Japan and Europe while remaining at a relatively low level in the United States. To help inform decisions about future activities in the United States, the EAR Program engaged researchers from PATH to conduct a study of vehicle-highway automation research and development activities in Europe and Asia.

The research involved a number of meetings, demonstrations, and site visits, as well as a literature review that looked at automation capabilities ranging from driver assistance to fully automated driving.

“Sponsorship for a lot of the development work in Europe and Japan is not coming out of transportation agencies; it’s coming out of the ministries and agencies assigned to promote industrial competitiveness and improve the economy,” says Shladover. “They are taking their obligations under the Kyoto Protocol very seriously, so reducing carbon dioxide emissions is driving this work.”

In Europe, perspectives differ between the organizations that approach cooperative vehicle-highway automation systems as automotive products and those that approach it as a means of improving public transportation. The former emphasize partial automation systems operating in mixed traffic, and the latter emphasize fully automated (driverless) vehicles in dedicated rights-of-way. Also, in addition to truck platooning, light passenger vehicle platooning has become a popular application for European study and development.

During testing in Japan, researchers placed this invehicle display screen near the passenger seat in a truck, so that they could see the status of the platoon system.

“Europe and Japan are conducting a number of studies on these automation systems right now,” says Robert Ferlis, technical director of FHWA’s Office of Operations Research and Development at the agency’s Turner-Fairbank Highway Research Center. “The applications that are commercially available are still limited, but research is being conducted to develop the next generation of systems.”

The EAR Program-funded researchers concluded that the United States can build on its extensive experience and capabilities in automated road transport, its current areas of international leadership, and the knowledge gained from programs overseas to advance a robust program in this field. Future actions could include conducting concept studies of automation applications to identify those that could be beneficial in addressing U.S. transportation problems.

Truck platooning is an area where the United States could pursue more indepth work to complement current activities in Japan and Europe. This research might include refining concepts of operation for truck platooning systems to address both urban and intercity applications, experimental work to develop reliable quantitative data about fuel and emissions-saving potential, and benefit-cost studies to support the definition of business cases for further development and deployment.

This four-truck platoon is driving on a German autobahn with 33-foot (10-meter) gaps between the vehicles. The test was part of a project sponsored by Germany’s Federal Ministry of Economics and Technology.

Advanced Traffic Signal Control Algorithms

About 8 percent of traffic delays on major roadways are caused by poor timing of traffic signals, according to Temporary Losses of Highway Capacity and Impacts on Performance: Phase 2 (ORNL/TM-2004/209), a study published by the Oak Ridge National Laboratory in 2004. To help address this problem, the EAR Program sponsored a project on advanced algorithms to control traffic signals. For this project, the program is partnering with the BMW Group, Caltrans, and the University of California, Berkeley and Riverside.

“Adaptive control produces benefits over fixed-time control, but implementation in real-world driving scenarios makes widespread adoption very challenging,” says Alexander Skabardonis, a research engineer at the Institute of Transportation Studies and a program leader at California PATH. As traffic control signals are replaced, there may be an opportunity to move from fixed-time to adaptive control, which then could provide additional benefits through connectivity with future vehicles that have the technology. Today, however, adaptive control might not be the best option for every intersection in terms of the cost-benefit ratio.

The project researchers investigated how vehicles and infrastructure could work together to provide comprehensive, real-time information on traffic through signalized corridors. The researchers studied signal control concepts from the standpoint of mobility, the environment, and safety.

Regarding the mobility aspect, the team explored strategies to avoid traffic saturation in the middle of a signalized corridor. The research involved the use of measurement samples to predict traffic, the average travel times, and the proportion of stopped vehicles. The researchers developed schemes such as gating to prevent queues from forming at the beginning or end of a signalized corridor.

For the environmental concept, the team looked at fuel consumption and emissions impacts, and how signal phases and timing can be used to save fuel and reduce air pollution. The researchers were able to quantify the savings and use that information to design optimal speed trajectories to enhance savings as vehicles drive through signalized corridors.

Shown here are several information displays and a reflector on the rear of a tractor-trailer truck traveling in a platoon in Japan.

Regarding the safety concept, they examined how to take full advantage of connected-vehicle data to analyze intersection geometry and detect approaching and waiting vehicles that might conflict. They studied how a system could give priority at intersections to emergency and transit vehicles, or, under low-traffic conditions, individual vehicles, thus reducing delays and the need to stop.

The private sector partner focused its research on developing algorithm sequences so that connected vehicles could inform only relevant traffic signals about the vehicles’ proximity, velocity, and signal request. In the study, information was sent from a traffic signal to a cloud-based (online) data center, and then communicated over a third- or fourth-generation network to invehicle applications.

Through this communication, vehicles would be able to display signal phase and timing information to drivers. Then, if necessary, the system could adapt each vehicle’s cruise control in real-time according to the vehicles’ trajectories. The goal is to move through a signal corridor without stopping. Known as Smart Cruising, the technology would enable a driver to choose between reducing travel time and increasing fuel efficiency.

Another technology, called “motor start stop automatically,” was developed by the private sector partner several years ago. The technology automatically shuts down a vehicle’s internal combustion engine when it is not needed. With this technology, a vehicle can sail along the road with the engine turned off, reducing fuel consumption and emissions. The technology was optimized by using advanced traffic signal control algorithms to determine the remaining real time at a signal in order to decide whether it makes sense to stop the engine.

The researchers found that basing traffic signal control on a more complete knowledge of the state of the traffic network could bring major improvements to the operational efficiency of traffic signals on arterial systems.

Researchers are using results from connected-vehicle tests to explore new ways to time traffic signals and develop control strategies for signals, which could improve older intersections, such as this one in McLean, VA.

Steps Ahead

Given the potential benefits demonstrated in these EAR Program-sponsored studies, the researchers are optimistic about moving forward with advanced solutions that can increase mobility and enhance safety.

“In fact, FHWA is committed to transitioning the results and findings into viable applications and further research projects,” says David Yang, who leads the Human Factors Team in FHWA’s Office of Safety Research and Development.

Yang provides an example: “Our traffic signal control research will pave the way for many future works to be carried out by the U.S. Department of Transportation [USDOT] and its partners to develop a new generation of strategies. Ultimately, this could lead to reductions in travel time and the frequency and duration of stops, in addition to reductions in pollutant emissions and fuel consumption.”

The next step of the signal control research team is to use real-world traffic and vehicle data to calibrate a traffic simulation that can represent a wide range of intersection configurations and traffic patterns. Findings are expected to demonstrate how control strategies based on data collected by vehicle probes and infrastructure-based sensors can effectively improve travel times, stopping frequency, fuel consumption, and emissions.

Similarly, the PATH research team is continuing to advance CACC applications developed under the EAR Program’s project on selected mobility applications. For example, an automobile manufacturer has funded a follow-on project to develop a second generation CACC system.

“Under the EAR Program, we used two CACC cars,” says Shladover. “With this follow-on project, we are using four cars equipped with CACC to show the differences in car-following stability more dramatically. Several car manufacturers are looking at this technology.”

Findings from the EAR Program’s Freeway Merge Assistance project have progressed to the field for prototype testing under a cooperative vehicle/infrastructure project at the University Transportation Center project funded by USDOT.

“We are developing the prototype algorithms, and we will be conducting the field testing at the Smart Road facility in Blacksburg, VA,” says Park. The Smart Road facility is a 1.7-mile (2.7-kilometer)-long, two-lane highway that is used for transportation research. The researchers recruited study participants to drive connected vehicles at the facility.

“We also have roadside equipment there,” adds Park. “The participants will drive the onramp vehicle, and our research team will drive two mainline vehicles. We will provide an advisory message based on numerous traffic scenarios, and see how the participants react to this. The results from this prototype testing will help us better understand driver behavior, which will culminate in further enhancements in the future.”

In fact, the future is now.

David Kuehn is the manager of FHWA’s EAR Program. Previously, he served as a community planner for the FHWA Office of Planning, Environment, and Realty and as a planner in local government. Kuehn entered Federal service as a presidential management fellow after completing a master’s of public administration at the University of Southern California. He also holds a B.A. from the University of California, Irvine.

Mark Fitzgerald is a senior writer at Woodward Communications and teaches writing at the University of Maryland. Before joining Woodward, he was the editor of several trade magazines and worked at the American Society of Civil Engineers. He has a B.A. in English from Franklin & Marshall College and an M.F.A. in creative writing from George Mason University.

For more information, visit www.fhwa.dot.gov/advancedresearch or contact David Kuehn at 202–493–3414 or david.kuehn@dot.gov.