Is this the future of railroading?
Autonomous freight cars could be just around the corner
The three startup companies developing autonomous, battery-electric, self-propelled freight cars all imagine a future where railroads see massive growth. They want to target short-haul freight where trucks dominate and railroads aren’t effective competitors. And they say they can solve the rail industry’s growth problem by using cutting-edge technology to upend the way railroads have operated for nearly two centuries.
The companies — Parallel Systems, Intramotev, and Glid — are led by former aerospace engineers who saw the same opportunity to disrupt freight transportation: Add autonomy to the efficiency of steel wheels rolling on steel rails, and you’re able to go after freight moving in lower volumes over shorter distances. This, they say, will enable a major shift of freight from highways to zero-emissions rail service, which will bring environmental, traffic, and safety benefits.
The three companies are taking slightly different approaches.
Parallel Systems aims squarely at short-haul intermodal. Its high-tech flatcars carry containers and by design would operate alongside conventional rail equipment, not with it. Parallel and two Genesee & Wyoming short lines are seeking Federal Railroad Administration permission to begin tests in Georgia this year.
Intramotev targets carload freight by retrofitting conventional equipment with autonomous, battery-electric propulsion systems that also can be operated by remote control. A handful of the company’s TugVolt hopper cars are in revenue test service on a pair of isolated mine railroads that are beyond the reach of federal regulators. The goal is to prove the technology before making the leap to short lines and Class I railroads.
Glid’s sleek, bimodal Gliders are on the drawing board, with the first hybrid prototype scheduled to debut this year. Gliders will be able to operate on rail or roads and are designed to handle any trailer, from dump
trucks to dry vans, which would open railroads to virtually the entire trucking universe.
What the three companies are proposing is a far cry from the automation being pursued by locomotive manufacturers Wabtec and ProgressRail and tech startups like Railspire. Instead of merely automating what already exists — like what mining company Rio Tinto has done in Australia with its heavy ore unit trains — the upstarts envision technology that will create an entirely new operating
model that should provide more frequent, more flexible, and more reliable service.
The railroad visionaries advising Parallel, Intramotev, and Glid see huge potential in the technology. It will, they say, finally allow railroads to really grow after decades of marketshare loss to highways.
The skeptics — and there are many — point to a host of potential hurdles. Will the technology work? Will it be safe? Will regulators give their blessing? Will rail labor unions approve? Will the public welcome trains rolling through neighborhoods and grade crossings with no one aboard? And will railroads shun innovations that they didn’t invent?
Although none of the vehicles would require engineers or conductors, Parallel, Intramotev, and Glid say their autonomous systems would ultimately lead to growth in overall rail employment. Put more freight on rails, they say, and you’ll need more people in terminals, in new office jobs, and to maintain and repair equipment, tracks, and signal systems.
In the next few years we’ll know whether the technology is a flash in the pan — or the spark that will ignite a new railroad revolution.
Parallel Systems
Los Angeles-based Parallel Systems, founded by a team of former SpaceX engineers in 2020, set out to find the answer to a single question: How can more freight be shifted from road to rail? They wondered what a railroad would look like if it were invented with the technology of today. This ultimately led Parallel to take a clean-sheet approach to railroad equipment and operations.
“Why does a train still need to be in its current form? Our process was to peel back the layers and try to get the fundamentals,” says CEO Matt Soule, one of the company’s founders.
Parallel’s engineers looked at advancements in self-driving technology and in batteries, then aimed to understand how a self-propelled, autonomous railcar could use existing infrastructure and operate alongside conventional equipment while helping railroads crack the short-haul market.
“The challenge is the massive scale that railroads tend to operate at,” Soule says. Moving high volumes of freight long distances — whether in bulk unit trains or 3-mile long stack trains — produces fat profit margins for railroads, which generally ignore moves of less than 500 miles. But that strategy also limits the markets where railroads can compete and the types of freight they can carry.
Enter the autonomous, battery-electric freight car. “It’s the autonomy that allows us to still accomplish truck-competitive unit economics without requiring massive scale,” Soule says. “When you can relieve yourself from that requirement, then you can start to develop more flexible services and new approaches to terminals.”
Parallel’s second-generation prototype flatcar can carry one intermodal container 500 miles on a single battery charge. The batteries in each truck power the car’s traction motors. The autonomous system relies on cameras and sensors, as well as software in computers on board the car and at a data center. And electronic brakes can bring cars to a stop in 10 times less distance than conventional equipment with air brakes.
Just as important as what the cars have is what they don’t: couplers. Instead, they’re equipped with bumpers that allow cars to travel together in platoons that improve aerodynamics, reduce energy consumption, and make efficient use of rail network capacity. Platooning is automatic and can be done on the fly. The cars move close to each other and touch bumpers. Once contact is made, each vehicle maintains a set force with the one in front by regulating tractive effort.
“We’re not proposing to move single containers through the network, because that is not a good use of network capacity. Nor is it the most aerodynamic,” Soule says. “We think the ideal platoon size is between 10 and 50 vehicles.”
A 50-car platoon would be about a half-mile long, and is the practical limit due to train dynamics. Ten is the lower limit, Soule says, because that’s where the energy advantages of platooning begin.
The cars are so unlike conventional rolling stock that the G&W and Parallel waiver petition asks the FRA to make the equipment temporarily exempt from 30 separate car and locomotive regulations that cover everything from cabs and
sanders to brake valves and uncoupling levers.
These differences make them a tool for taking on truckers.
“We’ll never be as flexible as a truck. We are absolutely not approaching this from the standpoint that trucks are going to go away,” Soule says. “But what we’re doing is working to enable modal shift. There’s a tremendous number of miles that are driven by trucks each year, and we think up to 50 billion miles of that activity could be shifted to rail with a platform like this.”
Parallel’s equipment would enable railroads to use small intermodal terminals where platoons would frequently arrive and depart. The idea behind Parallel’s terminals is to keep freight moving. And since they’re small, they could be located much closer to customers — or even at an auto assembly plant, warehouse, or steel mill.
“It’s all about asset turns. That’s the business case for autonomy. The equipment is going to be more expensive than conventional rolling stock. But because we’re not limited by availability of power or crew, we can move those assets a lot more than they do today,” Soule says.
Intermodal containers could ride Parallel’s autonomous cars from ports to big box retailers’ warehouses and transload centers, eliminating the trucking dray move between the ports and the customer. Or they could deliver international containers directly to inland markets within 500 miles of the docks: Think Long Beach, Calif., to Phoenix; Savannah, Ga., to Atlanta; or Newark, N.J., to Buffalo, Pittsburgh, or Cleveland.
For domestic business, Parallel’s cars could be used to connect small markets on either side of the Mississippi River, which aren’t well-served by Class I railroads today, as well as to eliminate crosstown drays in Chicago. They also could serve as a feeder network that connects existing double-stack intermodal terminals in Chicago with destinations like Milwaukee, the Twin Cities, or Detroit. Or they could be used to launch new service between cities like Denver and Salt Lake City.
“This is a real opportunity for the industry that helps with growth but also helps to strengthen and make more resilient supply chains,” Soule says. “A lot of our strategy is to improve the quality of service for rail … This platform has strong abilities to help do that.”
Intramotev
Boeing engineer Tim Luchini was developing flying cars and package-delivery drones when his business student friend, Alex Peiffer, suggested a more down-to-earth project: Autonomous freight cars.
Peiffer was researching how emerging technologies could shape logistics and supply chains as part of his MBA work at the University of Southern California. What would happen, he wondered, if a delivery drone’s sensor package was applied to trains? The duo’s first concept was underwhelming. “But we riffed off of it,” Luchini says. “It got to the point where we said, ‘Hey, every railcar could drive itself, so why don’t we try that?’”
And so St. Louis-based Intramotev was born in 2019, with Luchini as chief executive and Peiffer as chief operating officer. Their goal is to apply an engineering solution to a railroad business problem.
“How do you drive volume to rails today when people want the speed and flexibility of trucks but they want the cost advantage and they want the
environmental advantage of rail? How do you marry those two together? They’re conflicting, it seems,” Luchini says. “You have the lowest-cost mode of transportation and you have the fastest mode of transportation over land. And we’re really trying to build the best of both worlds: Something that has the speed and flexibility of a truck, takes advantage of the rail system, and ultimately leans into what the customers want, which is just in time delivery.”
The result is the TugVolt, a battery-powered freight car that can be fully autonomous, operated by remote control using a smartphone app, or tucked into a traditional train.
Adding the Intramotev tech to conventional freight cars makes the TugVolt fully compatible with existing operations. As part of the freight car retrofit, Intramotev adds a secondary electronic braking system that’s triggered when the car’s sensors detect an obstacle ahead, such as at a grade crossing. The stopping distance is dramatically reduced compared with conventional pneumatic brakes.
Intramotev envisions three modes of TugVolt operation.
First, it can run in a train just like any other freight car, making them easier for railroads to adopt.
The second is point-to-point operation — like from California to Chicago — that combines conventional operation for the long-haul and autonomous or remote control operation for short-haul regional and local delivery. “The reality is that stuff ’s not all consumed in Chicago,” Luchini says. “You’ve got to get it to Gary, Ind., you’ve got to get it to Milwaukee, you’ve got to get it to Peoria, Ill. So you can cut that car off … and let it go.”
The third is using TugVolt like an autonomous or remotecontrol locomotive in local service. “It can pull five fully loaded cars with it. So that’s really where you get the unit economics of a big train, but on a small train more in line with what the customers … are asking for,” Luchini says. A manufacturer that uses plastic pellets, for example, likely prefers to receive a handful of loaded cars per day, not unit-train quantities. TugVolt can bring down the cost of local service and make rail competitive with trucking in shorter hauls, Luchini says.
Class I railroad CEOs say the key to volume growth is providing better service. “But they don’t have the tools to do that,” Luchini says. “This is that tool.”
The base model TugVolt is equipped with 100 kilowatthour batteries — about the size of a Tesla battery pack — that can move 100 tons of freight 100 miles on a single charge. “You’re getting 3.5 times the weight of a semi a hundred miles,” Luchini says. A longerrange version can run 600 miles between charges. The batteries can be charged in as little as 30 minutes.
Intramotev is taking a threephase approach to introducing the TugVolt. The first target: Mines, ports, and intraplant railroads not regulated by the Federal Railroad Administration. This allows Intramotev to test, refine, and prove the technology with revenue customers. The first two customers are isolated mine operations, including the Cumberland Mine Railroad in Pennsylvania. Cumberland began using its first TugVolts in late December.
Mining operations are good testbeds, Luchini says, because they move millions of tons of material per year, tend to be out of sight, have labor challenges due to their remote locations, and want to electrify their operations for environmental reasons. TugVolt hoppers can reduce operating costs by 30 to 80%, with a payback period measured in months. “It’s a no brainer from a business case,” Luchini says. “You get huge cost savings and you electrify.”
The second target, likely three to 10 years away, is common-carrier regional and shortline railroads, once the system is proven to be safe and reliable, regulatory hurdles are cleared, and the public is comfortable with the concept of autonomous rail vehicles. The final market, likely more than a decade away, would be Class I railroads. In either case, FRA waivers would be required to permit autonomous and remote-control operation.
Any freight car can be retrofitted with Intramotev’s tech, not just the hoppers destined for mine use. “Intermodal is
really interesting. We have that carbody type already,” Luchini says. “Our general approach is to basically be scalable around any car type.”
Gliders
Glid’s idea of a better mousetrap is one sleek machine. The company’s Glider system pairs autonomous, battery-electric front and rear bogies that can carry a fully loaded semi-trailer. The bogies use rubber tires when on the road. Pull the Glider onto a track — at a grade crossing, a paved lot, or even just gravel spread up to the rails — and it will deploy flanged steel wheels. The rubber tires retract. And then it becomes something akin to a self-propelled RoadRailer. “We can transport any commodity you put on rail and road,” says CEO Kevin Damoa, a former SpaceX employee who is both a mechanical engineer and a battery engineer.
To avoid regulations on highway weight standards and autonomous operations, Damoa envisions Gliders operating on private roads, such as at industrial parks, that are close to the tracks. Truckers would use conventional rigs to transport the trailers on public roads, just like today’s intermodal dray operations.
Damoa says Glid’s initial target market is short lines. They tend to operate older, less efficient, more polluting locomotives that are costly to operate and whose emissions disproportionately affect socioeconomically disadvantaged residents who live on the proverbial wrong side of the tracks.
The first prototype, which will be a hybrid that features a biodiesel engine to charge the batteries and extend their range, is being built at a factory in Stockton, Calif. The conceptual design looks a bit like a futuristic hot rod convertible. “We’re trying to make it as close to that as possible,” Damoa says. “When I designed it I said, ‘Hey, I need it to be sexy.’”
It will be tested on private trackage in Vancouver, Wash., that connects with short line Portland Vancouver Junction Railroad. “We’ll prove out the operation,” Damoa says, and then take the lessons learned and apply them to a 2.0 prototype. A Vancouver trucking outfit, Taylor Transportation, has ordered two prototypes.
Ultimately Glid wants to more fully test the prototypes on the 33-mile PVJR, which handles lumber, aggregates, glass, and sand — some of which could shift to local Glider moves that would avoid local transloads, Damoa says.
Glid, a unit of Salt Lake City-based Genesis Electronics, aims to use a different business model: Gliders as a service offered to short lines and truckers. The company will compete like a freight broker, and shippers can pay for one-time moves or pony up $15,000 for a monthly subscription per Glider, plus a certain cost per ton mile. Glid will maintain the vehicles.
Operations opportunities
Experienced railroaders advising the three startup companies say autonomous railcars could open up a wide range of significant operational changes in a tradition-bound industry. No, the autonomous equipment won’t replace the bulk unit train or the 12,000-foot, longhaul double-stack train.
But the self-propelled equipment can enable fast, frequent service. There would be no more waiting for a lone daily train or local service provided just three days per week. And there would be no lingering in a hump yard or intermodal terminal for a day, no need to run a 150-car merchandise train, or to wait for a rested crew or a late inbound train’s locomotives.
Instead, cars could release right after they’re loaded or unloaded in what Dean Wise, a consultant and former BNSF Railway executive who is a Parallel advisor, calls an “operational ballet.”
Terminal operations would be like going to the McDonald’s drive-through: Within 2 minutes, a container could be lifted onto a waiting Parallel car and be on its way. The cars could gather at a nearby siding or yard, classify themselves into platoons, and then head to their destination. What they will not be, Wise says, “is ants marching all over the network.”
Concerns about short, autonomous platoons gobbling up more than their share of mainline capacity are overblown, says Marty Schlenker, a former BNSF and CSX executive who is Parallel’s head of strategy.
In fact, Parallel cars — and therefore platoons — will have a higher power-to-weight ratio than conventional locomotive-powered trains. “They will behave more like light engines than trains,” Schlenker says. “Picture yourself as a crew member on the second move out of a terminal. The dispatcher comes on the radio. Would you rather hear, ‘You’re next after the stack train’ or ‘You’re next after the light engine’?”
When a Parallel platoon runs up on a conventional train out on the main line, it will be capable of entering the preceding train’s signal block at restricted speed, just as a conventional train would. “Its short length will mean that it promptly gets out of the way of any train following it,” Schlenker says. “And shorter length can be beneficial when a platoon follows a conventional train on single track and an opposing platoon or train must meet.”
In some instances a platoon could take up less capacity than a conventional train. If an 8,000foot train pulls into a 10,000foot passing siding, a Parallel platoon could pull in behind to await a meet — something a traditional freight train could not do, Schlenker explains. In addition, tucking a train and a platoon into a single siding also would free up the closest passing siding behind them.
Advancements like moving block technology — in which multiple trains or platoons can occupy the same fixed signal block under directional authority at speeds above restricted speed — would enhance operations and increase capacity. “Even with current signaling technology, Parallel’s platoons will have a light footprint, and we are developing ways for them to have an even lighter touch on the network in the future,” Schlenker says.
Cathcart Rail CEO Casey Cathcart, whose company operates three short lines and 18 railcar repair facilities, is an advisor to Intramotev. He says Intramotev equipment would dramatically improve the way short lines operate and interchange with their Class I partners.
At Cathcart’s Belpre Industrial Parkersburg Railroad, CSX delivers inbound interchange cars overnight to the short line’s yard in Parkersburg, W.Va. The next morning BIP switches the inbound cars, rolls them across the Ohio River bridge, and distributes them to customers in Ohio.
With Intramotev cars, CSX could deliver an interchange train and the cars would release themselves one by one and run directly to the customer’s gates. They’d be waiting for the customers when their facilities open in the morning, and plant employees would operate the cars on intraplant trackage. The process would work in reverse for interchange to CSX: The cars would head from customer facilities and gather themselves in the yard in Parkersburg.
“Not only have you streamlined shortline operations, you’ve streamlined in-plant, captive-use cases and you … can do away with all the extra yard trackage. You don’t need it. Now all you need is a main line and maybe a siding to peel out all the non-Intramotev cars,” Cathcart says.
A short line could run fewer switch jobs and condense work into fewer shifts, Cathcart says. “You will have a lot more people behind screens than you have down in the yard,” which he says would improve employee safety.
The lower-cost operation also would open up more short-haul moves to rail.
“The dispatcher could actually be Intramotev, with a single person sitting in a dispatch office overseeing that move versus my own people,” Cathcart says. “The cost for that is probably substantially cheaper than having a yard superintendent and other people.”
Yet large portions of the North American rail network are dark territory with hand throw switches. To take advantage of automation, railroads or customers would have to automate switches and make other investments in order to make door-to-door operation feasible. And coupling and connecting brake hoses would need to be automated, too.
Optimism and skepticism
Industry observers say widespread adoption of autonomous technology faces long odds and a time frame measured in decades, not years.
“It is exciting to see the level of innovation that the rail industry is generating — in these types of new conveyance concepts as well as other areas like inspection technology, car telematics, and in-cab automation,” says Adriene Bailey, a partner at consulting firm
Oliver Wyman. “There may be closed-system applications for these that will make sense in the short term.
“The challenge is and will continue to be the need for standards and interoperability across the rail networks,” Bailey says. “In addition, the vast, expensive, and long-lived installed asset base will make any transition to a new ‘standard’ of conveyance a complex and most likely multi-decade initiative. It’s super important to get it right when thinking about ‘the next gen’ of railroading — because false starts will be very expensive.”
Todd Tranausky, former vice president of rail and intermodal at freight forecasting firm FTR Transportation Intelligence, says there are too many hurdles stacked against autonomous freight cars for them to enter service anytime soon, if ever.
“It’s great if it can be made to work. But I am skeptical,” he says. “This is the industry that can’t even agree on a single railcar visibility solution. Now we’re going to get railroads to agree, labor to agree, shippers to agree, and regulators to agree on a new way of moving cars around? It feels like a massive leap.”
The FRA accepted the G&W waiver petition in January but sought written comments and said it may hold a public hearing on the Parallel test proposal.
Greg Regan, president of the Transportation Trades Department of the AFL-CIO, says the potential impact of autonomous technology poses more questions than answers, particularly on safety. “People jump to the rosiest, sunniest possibility without considering all of the really important things that go into a safe operation,” he says.
Labor unions would be willing to have discussions about autonomous operations as long as railroads view workers as partners, Regan says, noting that ultimately the unions must adapt to new technology as they have throughout railroading history. “We could have a conversation about how do we use this new technology to supplement — and not one that will supplant — existing operations,” Regan says. But he doubts that Class I railroads have any interest in adopting autonomous tech unless it means whacking jobs.
“If the railroads’ recent behavior would be any indication, then they would use it to eliminate headcount and lower their overhead as much as possible,” he says. “They’ve shown very little interest in actually growing and expanding their operations over the last five or six years.”
National Transportation Safety Board Chair Jennifer Homendy in January told Congress that she had safety concerns about autonomous trains based on lessons learned from investigations of crashes involving selfdriving cars. “I would not like to see a 2- or 3-mile-long train — much less a 4-mile train, which I once saw a consist for — with nobody on board,” she said.
Intermodal analyst Larry Gross says Parallel and Glid could be game-changers, if they can get past the regulatory, labor, and technical challenges.
“Either of these is a tool for getting at the rest of the market, which is THE market. Intermodal is an afterthought right now for shippers,” Gross says. “Either of these is a quantum jump toward adapting domestic intermodal to the needs of the domestic freight market.”
The current domestic intermodal system — high volumes moving long distances between major terminals — doesn’t fit how freight moves in shorter hauls to widely dispersed destinations. “This industry needs disruption,” Gross says.
Parallel could use smaller, less expensive, and more numerous terminals to speed freight, reach smaller markets, and be an effective feeder system for the long-haul intermodal network, Gross says. Gliders don’t need a terminal and could tap the much broader opportunity to handle highway trailers. “It’s very ingenious,” Gross says.
But he cautions that railroads like standardization — and intermodal innovations like RoadRailer and Expressway ultimately failed to survive after domestic double-stack reached critical mass.
Wise says there’s widespread interest in Parallel’s system. The company has had discussions with railroads, shippers, trucking companies, ports, and steamship lines. “No one threw us out of the room and said this is crazy,” Wise says.