The Trains Run on Time with AI: How Artificial Intelligence is Transforming the Rail Industry

The United States is making strides to catch up to other developed nations’ extensive rail networks. For the first time in over a century, the U.S. is investing heavily in passenger rail infrastructure. Nearly $370 billion in federal funding was announced from the Biden administration in 2022, for 46 projects in 31 states. Much like the U.S. did in the first transcontinental “rail race” of the 1890s, rail teams are working around the clock to deliver projects on time.??

However, the world is a vastly different place now than it was when the Union Pacific and Central Pacific Railroads first battled to see who would lay track across the country the fastest: all the expenses needed to complete rail projects, from labor to materials, are skyrocketing. Post-pandemic labor shortages and supply chain issues are holding back even the most organized and efficient project teams, leading to massive budget overruns and delays of months or even years.??

The rail industry has lost over 45,000 workers in the last 6 years and 96% of transit agencies are reporting shortages of labor. Supplies like lumber and steel are more expensive than they have been in recent memory; lumber prices are up 14% in 2023, while steel shortages have challenged manufacturers since 2020 with no signs of slowing down.??

How will the rail industry meet their goals with such limited resources? Many are turning to AI. ?

The Growth of AI in the Rail Industry?

Rail projects are highly visible, vulnerable to public and legislative backlash when they fail to debut on deadline or when already active lines are delayed or disrupted.??

In America’s new “rail race”, the stakes are higher and so are the costs. In this challenging environment, rail firms are turning to artificial intelligence to increase efficiency for construction, design, and operations.??

In fact, AI-related patents in the industry have dramatically increased in the last several years. According to Railway Technology, applications for AI patents by rail firms increased by over four times between 2021 and 2022 alone.? Many of the world’s top rail and engineering firms are employing AI solutions, including AECOM and HNTB. ?

Artificial intelligence shows promise for equipment maintenance, track monitoring, driverless freight trains, subsurface imaging, and more. We’ll examine several examples of AI in the rail industry below.??

Monitoring Tracks?

Britain’s Network Rail, which owns and operates the country’s rail infrastructure, is using AI technology?to monitor more than 20,000 miles of track across the country.???

Network Rail is employing a phone-sized device which can be attached to the front of any train and capture raw video footage of the track. Then, AI algorithms analyze the footage in real time and alert Network Rail of anomalies that could indicate weak areas or possible damages.?

?The company then sends out preventive maintenance teams to repair the track before it delays any trains or causes safety risks. By using AI, Network Rail can analyze vast amounts of data about their rail network quickly and accurately.??

Repairing and Maintaining Equipment??

Equipment failure on trains is frustrating at best and incredibly dangerous at worst. To avoid delays and keep passengers safe, many rail firms are turning to “predictive maintenance”: using AI to determine when equipment will fail so that it can be replaced before it becomes a liability.??

By employing AI algorithms, rail companies can analyze billions of data points from equipment sensors to pinpoint future failures, predicting engine problems weeks in advance.???

Driverless Freight Trains ?

Self-driving cars are on the road in several American cities, and someday soon driverless freight trains might be on the tracks. ?Completely electric, self-driving freight train cars are currently under development, capable of carrying single or double-stacked standard shipping containers.??

Because they will operate autonomously, once fully developed and tested, self-driving freight cars will be “self-platooning”, meaning that they could connect to one another and travel as one long train, or split off from one another and go to different routes autonomously.??

Accurate Subsurface Imaging???

Utility strikes and redesigns due to inaccurate underground records and maps frequently plague the rail industry. Protracted utility relocation recently delayed Maryland’s Purple Line project in the suburbs of Washington, D.C. by an additional seven months. The line was originally slated to open for riders in fall 2026 and is now expected to open in mid-2027 at the earliest.??

Exodigo offers a solution to deliver accurate, complete, and comprehensive underground mapping, by processing data much faster and more accurately than human beings alone. With the help of AI, Exodigo can synthesize, analyze, and process over 500GB of data per acre in merely days, creating an intuitive visual output that can be employed in the field without the frequent inaccuracies present in (often out-of-date) written records.??

With the use of Exodigo’s technology, it is possible that the delays in Maryland could have been avoided or mitigated significantly. Exodigo can find and identify 20-50% more lines than alternatives, according to Common Ground Alliance’s 2022 Technology Report.?

By integrating data from advanced sensors across different fields of physics, we generate a comprehensive, non-intrusive representation of the underground environment. Armed with precise, complete knowledge, project stakeholders can expedite project delivery, leading to compounding returns on investment through significant time and cost savings, enhanced safety measures, and improved sustainability.??

The Power of AI for the Rail Industry??

AI is a clear pathway forward for rail firms navigating escalating costs, supply chain disruptions, labor shortages, and even climate change. By increasing accuracy and efficiency, AI is keeping passengers safer, shortening project delivery timelines, and reducing the carbon footprint of the industry.??

We look forward to seeing how developments in AI continue to serve the rail industry and partnering with forward-thinking rail firms seeking innovative solutions.??