AI Company Aids North American Diesel Trains in Reducing Pollution

Rail transport has received relatively little attention in terms of reducing its climate impact, despite the fact that transportation is accountable for a significant portion of global greenhouse gas emissions. While electric vehicles and sustainable aviation fuel have been promoted as means to decrease emissions from air and road travel, the focus on mitigating the environmental effects of rail travel has been limited.

A Canadian startup wants to reduce this effect. Montreal-based RailVision Analytics has developed AI-enabled software that helps locomotive engineers make small changes to how a train runs, which can lead to big savings in diesel fuel. This could help freight and passenger trains cut the roughly 100 million tons of planet-warming gases released into the atmosphere each year.

“It’s just like Google Maps,” says Dev Jain, founder of RailVision Analytics, of his AI app, which can be downloaded to a tablet and used offline. Google Maps tells drivers to turn right or left, while RailVision’s app directs locomotive engineers to “stay idle” for the next mile or “increase speed.”

Basically, the idea is to allow locomotive engineers to eliminate pointless practices and tap into unseen forces that could help their driving. If you’ve driven a car, you’re almost certainly familiar with the law of inertia discovered by Isaac Newton in 1686, which states that a moving object can continue in the same direction unless acted upon by a force. This means that the vehicle can stay in motion – that is, coast – without human-made propulsion power until friction or other forces stop it.

But “riding the train is like riding a roller coaster,” says Jain. The length of the trains means that when trains run on tracks with varying elevations, often some train cars start to climb uphill, while others are still downhill. This poses a difficult question for locomotive engineers: Should they coast or throttle to get more horsepower?

To solve these puzzles, RailVision collects data from train traffic to simulate train dynamics and uses an algorithm to determine the most fuel-efficient way to drive. While the AI solution is still in its infancy, Jain says it has helped Metrolinx, a government agency serving millions of riders in the province of Ontario, save more than 1.5 million liters of diesel fuel during a year-long trial. RailVision also counts Via Rail Canada Inc., the country’s largest long-distance passenger rail operator, as an early adopter.

Trains are probably the most climate-friendly way to get around. According to the International Energy Agency, train journeys release an average of 22 grams of carbon dioxide per passenger kilometer worldwide, which is lower than air travel’s 123 grams and passenger car journeys’ 145 grams per passenger kilometer.

However, diesel is a particularly dirty fuel. Unlike Europe and Asia, where electrification has advanced in passenger rail, most long-distance trains in the United States and Canada run on diesel fuel — a status that is unlikely to change anytime soon, says Bruno Idini, a transportation analyst. IEA.

This is because running trains with electricity requires an expensive infrastructure upgrade, and the region’s relatively scarce passenger traffic cannot justify this investment, says Idini. It is also expensive to build electric rail networks for freight trains to bring goods to towns and villages that are spread over vast distances. While some rail companies have begun to replace diesel with lower-emissions biofuel, that replacement is still “a drop in the bucket,” he adds, citing its high cost and limited supply, among other constraints. Artificial intelligence could offer a relatively inexpensive alternative to cutting the carbon footprint of rail transport. “It was the first time we saw an innovation that was easy to install and apply to our reality,” says Francoise Granda Desjardins, Via Rail’s sustainability advisor. which has been testing RailVision’s AI solution since 2021.

Via Rail started its experiment with a simulation test, during which engineers sat in a replica of a locomotive cab to drive trains in virtual reality. After six months, the engineers used 15% less diesel under the AI-generated guidance than when driving without it.

The company has since tested the technology on passenger trains that run between Toronto and Ottawa. Although Via Rail has yet to finalize the findings, “it’s been a positive experience so far,” says Desjardins.

Founded in 2019, RailVision raised $4 million in a seed round last year from investors including Trucks Venture Capital and Blackhorn Ventures. Since fuel is the second largest operating cost for North American railroads after labor costs, the startup’s solution is turning heads. Genesee and Wyoming Inc., a multinational rail company whose business includes freight trains in Canada, recently launched a pilot with RailVision on its freight service between Montreal and Quebec City.

RailVision charges railway companies a subscription fee for its artificial intelligence control. While Jain declined to disclose pricing, he says the service “usually pays for itself in a few months” in fuel costs saved.

Climate benefits can also be considerable. Take Via Rail. If the artificial intelligence solution were implemented in all its trains, the expected 15 percent fuel savings would reduce carbon dioxide emissions by more than 20,000 tons per year, the company estimates. This corresponds to more than 4,000 cars being taken off the road per year.

However, some wrinkles need to be ironed out in order to realize this potential. Because driving in the real world is more complex than in a controlled environment, there is a difference in performance between simulated and real train operations, according to Desjardins. In addition, fuel is only saved if people follow the advice of artificial intelligence. As safety regulations prevent locomotive engineers from looking at the app or listening to audio alerts while the train is in motion, they are asked to memorize all instructions before departure. It remains to be seen whether locomotive engineers are ready to improve fuel efficiency in their daily work, or whether they will rely primarily on artificial intelligence.

In order to achieve zero emissions, rail transport must eventually give up fossil fuels. But until that happens, “technologies like RailVision are really useful,” says Desjardins. “If we reduce fuel by even 15%, that’s significant.”