The National Laboratory of the Rockies (NLR) is using artificial intelligence, machine learning and data analysis to address transportation challenges, according to nlr.gov. Its projects cover urban traffic management, connected vehicles, airport shuttle operations, truck efficiency and public transit research.
In a Chattanooga, Tennessee, pilot, researchers tested an adaptive traffic signal algorithm developed using real-world data and supercomputer simulations. The system reduced traffic delays by 32% and increased fuel savings by 27%, the laboratory reported. Researchers have also developed models that dramatically accelerate simulations involving 4,000 connected vehicles.
Working with Dallas Fort Worth International Airport, NLR developed a simulation tool to improve shuttle scheduling and routing. The system can model a full day of operations in two seconds with 98% accuracy. Optimized schedules and a modernized fleet helped the airport save more than $1 million during the project's first two years.
Other research uses computer vision to examine highway conditions affecting truck platooning, in which electronically linked trucks travel closely together to reduce aerodynamic drag. Researchers found that roadworks, traffic slowdowns and other obstacles help explain why real-world fuel savings are lower than those observed in controlled tests.
NLR also applies large language models to social media analysis through its SAMBA tool, which helps researchers assess public views on transit, airport mobility and automated vehicles. The approach can produce preliminary findings within minutes, while still requiring human review to verify results and interpret them correctly.