Multiyear NREL Study Uses Digital Twin of DFW International Airport To Inform Long-Term Investments at Major Transportation Hubs

Shuttle Bus Route Optimization and Expanded Parking Choices Could Reduce Energy Use and Operational Costs

July 21, 2021 | Contact media relations

Photo outside of Dallas-Fort Worth International Airport, where passengers unload and load into cars.
As demand for air travel continues to increase, airports will face challenges with more passenger vehicle traffic. NREL scientists are using a "digital twin" to understand how different policies and technologies could support more-efficient ground access travel. Photo by Dennis Schroeder, NREL

Rapidly emerging smart mobility technologies could dramatically decrease energy consumption and operational costs at airports, but how do operations staff choose the right innovations and policies to adopt?

Researchers at the National Renewable Energy Laboratory (NREL) and Oak Ridge National Laboratory are providing complex transportation hubs with decision support and actionable insight through Athena, a collaborative multiyear research effort initiated in 2018 to understand how airports can integrate and adapt to transformative technologies.

Funded by the U.S. Department of Energy Vehicle Technologies Office and in partnership with Dallas-Fort Worth International (DFW) Airport and industry, Athena leverages the powerful scientific computing capabilities at the Energy Department national laboratories to model current and future scenarios at DFW—one of the world's largest airports to achieve carbon-neutral status—as it continues to seek new ways to reduce carbon emissions.

Now in the third year of the study, the Athena team has published a series of three journal articles sharing the key results from their work.

"We've designed these tools and approaches to be generalizable so that Athena models, methodologies, and findings can inform long-term investments at U.S. airports and seaports over the coming decades," said Caleb Phillips, NREL senior scientist and Athena project lead.

Using Data-Driven Models and Artificial Intelligence To Improve Airport Operations

In the first phase of the study, the Athena team determined the best approach to forecast traffic demand and create high-fidelity simulations of traffic inside the airport. They used two years of detailed data for individual vehicle arrivals and departures, aircraft movements, and weather at DFW to evaluate multiple traffic prediction and simulation methods.

Athena researchers determined that several of the prediction models worked well when forecasting traffic in the next 30 minutes, including the classic time-series forecasting approaches favored in biological or finance applications—an approach called SARIMA (Seasonal Auto Regressive Integrated Method)—as well as modern deep learning methods. For longer prediction horizons, the machine learning and artificial intelligence methods offered the best predictions.

To model short and long time periods, Athena combined a demand forecast with a traffic microsimulation framework to produce a first-of-its-kind, data-driven operational model. The resulting operational intelligence tool, described in a just-published Journal of Air Transport Management article, provides a "digital twin" of the airport to simulate different scenarios and explore the impacts of policy changes and infrastructure expansion.

"We used the digital twin to study how different traffic control strategies influence congestion and energy consumption," said Monte Lunacek, Athena scientist and lead author of the article. "We found that when simulated passengers are notified of high traffic volume and go to a recommended alternative parking lot or terminal, it saves 34% of fuel consumption and nearly 5 minutes of travel time during the peak hour from 6 to 7 a.m. Without the digital twin, we wouldn't be able to understand the impacts of these type of changes."

Using the digital twin, the team also explored the impact of the adoption of emerging mobility technologies, such as autonomous vehicles. Simulation results showed that the use of battery-electric autonomous vehicles for the current portion of airport trips by passenger vehicle would lead to the highest reduction in fuel consumption and emissions at DFW. On the other hand, if every patron going to and from the airport switched to autonomous vehicles, the traffic volume to and from DFW would immediately surpass the airport's current capacity, causing significant increases in congestion, fuel consumption, and emissions. These results demonstrated the high cost of uncontrolled adoption of new technologies and the importance of planning for future mobility changes.

The team believes airports like DFW can use this digital twin to evaluate many operational scenarios, react to problems as they arise, and generally improve day-to-day procedures to maximize efficiency and reduce emissions and wasted time.

Getting People to and From Airports With a Lower Environmental Impact

Today, in addition to driving yourself and paying for parking, you have several ways to get to and from the airport, such as using Lyft, Uber, or taxis, getting dropped off by a family member or friend, and public transportation. With curb space becoming scarce and different modes of transportation eating into parking revenue, airport ground access travel impacts future congestion, revenue, and energy consumption. In the near future, electric vehicles and autonomous vehicles are expected to increasingly occupy this space as well.

In another phase of the study, Athena expanded on existing literature on airport ground access choice by estimating a joint model of airport access mode and parking product choice that considers today's diverse options for getting to the airport. This approach generates more accurate and realistic values of travel time and better predictive power for use in airport access mode decisions. Interesting trends emerged, as described in a Journal of the Transportation Research Board article. For example, people who have flights later in the evening often choose Uber or Lyft (but not taxis) instead of being dropped off by a family member or friend. And female travelers are less likely to use shared modes such as airport shuttles, a transportation network company (also known as ridesharing services), and taxis than male travelers.

The model proposed in this phase of the study can provide greater understanding of travel behaviors for policymakers. And the model enables the evaluation of the "what-if" scenarios in future infrastructure planning, for example, how passengers' access mode and parking choices change with the implementation of a terminal curbside access fee. This is a critical consideration especially as airports are working to reduce emissions and encourage greener modes of transport. In a related study, the team found that if even 5%–10% of passengers shift to cleaner options like public transit, it can significantly reduce emissions, avoid congestion in the central terminal area, and delay the need for costly infrastructure investments.

Optimizing Shuttle and Bus Routes To Support Energy Efficiency Goals

Airport shuttles are becoming increasingly necessary as U.S. airports grow, but shuttles account for significant energy usage and emissions at airports. To reduce this energy consumption, routes and schedules must be optimized without largely affecting passenger's travel experiences.

To identify areas that could be more energy efficient, Athena researchers used a data-driven approach to route and schedule optimization. They looked at demand-adaptive shuttle schedules that adjust based on demand fluctuations throughout the day and week. They considered new, more efficient shuttle bus routing options and the trade-offs between passenger wait time and energy consumption.

Opportunities for electrification and the economies of fleet electrification were also included in the study. The results, outlined in a Journal of Air Transport Management article, show that small changes to existing routes could lead to a 20%–40% energy reduction in shuttle operations with minimal effects on passenger wait times. With more moderate ride wait times, up to 50% energy savings could be achieved, resulting in significant savings in emissions and costs for the port operator. In another study, the team found there are significant opportunities for airports to adopt electric vehicle shuttles, which would even further reduce emissions while also providing a quieter and smoother ride for passengers.

How Do Athena Results Apply to Other Transportation Hubs?

Results from the three peer-reviewed articles are critical to helping understand the effects of changing landscapes at airports. And the methods from all three studies can be applied to other similar-sized international airports to help plan and mitigate traffic. In future years, the team hopes to perform similar research at other U.S. ports, including those on Athena's coast-to-coast advisory board that includes nine of the country's biggest airports. The Athena project can also be used to plan for burgeoning extensions to freight and cargo, as well as to understand environmental justice and sociodemographic equity implications of mobility changes.

"The Athena approach of applying modern data-science to airport operations and planning gives airports the flexibility to test the efficacy of solutions before they're actually applied," Phillips said. "We're discovering that reducing energy use and emissions doesn't have to hinder passenger experience or airport revenue. It's a critical time to reconsider how we invest in major U.S. infrastructure—working directly with major transportation hubs, like DFW, has allowed us to develop a careful, customized, and modern approach that can serve as a winning example nationwide."

Learn more about the Athena project and NREL's sustainable transportation and mobility research.

Tags: Computational Science