An American startup is on a mission to change rail freight forever. They want to introduce self-driving, autonomous freight cars. The idea is to make rail freight more flexible and cost-effective, and to get more freight on the rails. Let’s take a look at what the future of freight could look like.
A New Paradigm?
The traditional operation of the railways has not changed much since they were invented more than 200 years ago. There is an engine at the front pulling one or more cars. The cars are connected by couplers that transfer the tractive force generated by the locomotive. This arrangement has proven effective in many ways: You only need one locomotive to move many cars at once, which also means you only need one driver. It is energy efficient because the cars can move on steel, which has less friction than roads. And the cars can move in the slipstream of the cars in front of them-something that trucks, for example, cannot do, although there is some research into platooning.
The traditional approach of locomotives pulling a string of rail cars also has some drawbacks. One is what to do if not all the cars have the same origin or destination. Traditionally, this is solved by building classification yards, where switching locomotives break up trains and reassemble them into new trains. Not only is this a very labor-intensive process, it also takes a lot of time and significantly slows down rail freight transit times. In this respect, trucks on the highway have an advantage because they can move their load directly from source to destination, often much faster than rail.
Which brings us to a question: Wouldn’t it be great if every car could drive itself? If it could, we would no longer need classification yards. We would not even need locomotives and engineers.
This is the question that Parallel Systems, a startup based in Los Angeles, California, is trying to answer.
Parallel Systems is currently working on autonomous railroad cars. These cars can carry a standard container. Each car is equipped with battery storage, electric motors, and sensors for autonomous driving.
Each car can move on its own and, for example, travel to a customer’s siding on its own without needing an engine and an engineer to get it there. But the real trick is that these autonomous cars can form a train that travels together. It is not really a train because there is no mechanical coupling and no locomotive pulling it. Rather, it is a platoon in which the cars coordinate with each other to travel at the same speed. These platoons can form on their own and break up later. By forming platoons, more cars can run on the same track than if each car ran alone.
This video demonstrates the main principles of platooning in a computer animation.
The company is not talking about the maximum speed of the cars. To be compatible with existing rail traffic, they would have to reach speeds comparable to conventional trains on the same track. At least in the USA, where the maximum speed of freight trains is rather low, this should not be a big problem.
The developers promise a range of up to 500 miles and that their batteries can be recharged within an hour. This should allow for high utilization of the cars with relatively low downtime.
State of Development
Last week, Parallel System released a video showing test runs of their second-generation prototype. It demonstrates platooning for the first time in reality. Watch the whole thing, I found it really impressive!
Currently, the system is still in the prototyping phase. Real-life testing will begin in 2024 on short lines in Georgia, according to the magazine Trains.
In addition, there will be a cooperation between the Australian company Arc and Parallel Systems to use the new system in the future Westport in Western Australia. However, the press release does not mention a timeline.
The autonomous wagons envisioned by Parallel Systems certainly have the potential to revolutionize rail freight. On shorter distances, they could allow railroads to compete with trucking for container and intermodal shipments. And on longer distances, platooning could reduce the overall travel time for freight by eliminating the need for sorting in yards.
Whether this future actually unfolds will depend on many factors. A few technological questions come to mind: How well can autonomous railcars sense their environment? All sensors will have to be placed under the freight (similar to the Carfax concept), which is much lower than an engineer’s viewpoint. This will limit the maximum field of view – will this be a problem?
Will the system be reliable enough to operate fully autonomously in the harsh environment of a railroad? There are many pitfalls in fully autonomous operation, often unexpected ones, which brought down robot taxi operator Cruise. And that was in San Francisco, which has a fairly mild climate. How reliable can a train be in all weather conditions? What happens when snow or dirt covers the sensors?
I also wonder about the capacity of the batteries. The company promises that the batteries can be recharged in less than an hour. And they will want to do that to keep their equipment running at the highest possible speed, so charging will happen several times a day. This is much more than a consumer BEV will have to endure. How many charge/discharge cycles can the battery withstand?
Much of the success will also depend on the cost of a car. Each car will need a lot of extra equipment compared to a conventional car. It will need batteries, motors, sensors, computers, antennas, and most importantly, an autonomous braking system. Conventional brakes use compressed air supplied by the locomotive. For autonomous operation, each car will need its own brake. And these additional systems will also need to be maintained, with an upfront investment in maintenance facilities and ongoing costs for personnel.
All of this adds up to a significant price tag for a car. The question is whether railroads will be able to generate enough cost savings from the new system, or attract enough customers from conventional trucking, to offset those costs.
At this point, it is too early to tell from the information available whether these issues can be resolved. But from what we can see at the moment, the technology looks really intriguing, and I sincerely hope it makes the leap from the prototype stage to widespread use in the real world. It has the potential to give a real boost to an industry that is over 200 years old.