Researchers at Oak Ridge National Laboratory and National Renewable Energy Laboratory have partnered with Chattanooga on the use of simulations to determine controls that would deliver energy efficiency while also optimizing drivers’ travel time, speed and safety.
By digging into traffic congestion data for the area around Chattanooga, Tenn., the Energy Department aims to reduce vehicle energy consumption by up to 20% and recover as much as $100 billion in lost productivity over the next 10 years.
Researchers at DOE’s Oak Ridge National Laboratory (ORNL) and National Renewable Energy Laboratory (NREL) have partnered with Chattanooga on simulations to determine controls that would deliver energy efficiency while also optimizing drivers’ travel time, speed and safety.
To do this, the researchers have tapped 500 sensors -- including automated cameras, traffic signals, vehicle GPS devices, radar detectors and weather stations -- to build a digital twin of the city. The Tennessee Department of Transportation has sensors roughly every half a mile that collect data for volume and speed for each lane of traffic, providing up-to-the-minute situational awareness, according to Jibo Sanyal, group leader and senior scientist at ORNL’s Computational Urban Sciences Group.
The digital twin serves three purposes, he said: intelligence on what is currently happening, what might happen and what to do about it.
For instance, if an intersection gets backed up middays, the researchers can use the digital twin to look for ways to alleviate that congestions. They can experiment with connecting to a small set of traffic signals to test algorithms that control include signal programming, alternative routing, speed harmonization, ramp metering and dynamic speed limits.
“To be able to control something reasonably well, we need to know what’s happening and then be able to anticipate what may happen,” Sanyal said. “That’s where this aspect of being able to connect to the sensing infrastructure that they have in the city is of paramount importance.”
By pairing the data with machine learning, the researchers can estimate energy use and energy loss, pinpoint which systems are losing energy and why and model realistic alternatives to changes in conditions and controls, NREL officials said.
“We’re trying to essentially explore this unknown space, understand how to ingest all these disparate sensors, build a situational awareness platform through the digital twin that could become the basis of effective control,” ORNL’s Sustainable Transportation Program Director Rich Davies said. Many of the unknowns lie at the boundaries of all these systems, he said. “Basic cellular communications we can rely on, but how exactly you inform a car about its surrounding traffic situation, there’s not really a standard for that.”
To handle the computation-heavy analysis of the real-time data and how changes affect outcomes, the team uses the Eagle supercomputer at NREL. It can carry out 8 million-billion calculations per second, which lets the researchers complete the computations in hours, minutes or seconds.
“If they were running on a laptop, they would be running for months,” Sanyal said.
Transportation operators and engineers and city managers can understand the results of the analysis through a visualization tool the team built for presenting the data in charts, graphs and animations. “We are computing various performance metrics and measures and differences as we change the system, so that gives us a good idea of performance-based measure of improvements,” he said.
The project began in 2018 with a two-year award from DOE, and the team recently renewed it for two more years to increase the area of study to include parts of Georgia, freight vehicles and more functionality. Chattanooga Public Works and the Georgia Transportation Department joined the Chattanooga and Tennessee DOTs and the Chattanooga-Hamilton County/North Georgia Transportation Planning Organization in the new phase along with industry partners Covenant Transport, FreightWaves and Siemens Mobility.
DOE would ultimately like to see the researchers achieve “a 20% energy savings in the entire area,” Davies said. With the growth in connected and automated vehicles and increasingly intelligent transportation infrastructure, introducing more active controls will allow officials to optimize the traffic signals, routing and other infrastructure and take advantage of improvements in vehicle drive trains, he said.
Such an optimized system is not just around the corner, though. It will need support from local leaders.
“We were worried [about] how long it was going to take to get there because the level of connectivity is relatively small, the level of automation is very small,” Davies said. “We believe that the solution is probably going to take something that involves the people who operate the roads and have the jurisdictional authority to make changes into the system.”
The technology developed for Chattanooga is transferrable to other places and other uses, such as studying the interface between the electric grid and electric vehicles that need charging. “We believe this sort of situational awareness will help us do the planning around the charging in the future, knowing both the traffic counts and the grid situation,” Davies said.