There’s no question that autonomous trucks pose a threat to the railroad industry. They’re not here quite yet, but TuSimple is proving the concept with driverless rigs that are making revenue runs on highways in the Southwest. And other companies are working to develop battery-electric trucks.
If proven feasible and deployed in sufficient numbers, driverless electric trucks would not only provide better service than railroads. They’d be cheaper, too. This combination of potential developments prompted an investment firm to issue a dire prediction: The sun is setting on the railroad industry.
ARK Investment Management argues that widespread adoption of autonomous electric trucks will put railroads at risk of bankruptcy — or worse.
“We believe that during the next five years, autonomous electric trucks will commercialize and take share from rail operators with more cost-effective, door-to-door service,” ARK wrote in a 2020 white paper about the impact of disruptive technology. “If autonomous vehicles proliferate into various form factors, including flying drones and rolling sidewalk robots, we believe freight rail companies will have trouble competing with antiquated technology tied to dedicated infrastructure assets. ARK wonders which, if any, freight rail operators will survi
ARK’s report lumped railroads in with four other risky bets, including physical bank branches, brick-and-mortar retail, the auto industry, and providers of cable and satellite television. (As an aside, I haven’t set foot in my bank branch or the mall in years, and I cut the cord with my cable TV service a little over a year ago. So in my world three out of ARK’s five predictions seem spot on.)
Not everyone agrees. In his annual letter to shareholders, Berkshire Hathaway Chairman Warren Buffett took the opposite view while recalling what led his company to acquire BNSF Railway a dozen years ago. “And here I’ll venture a rare prediction: BNSF will be a key asset for Berkshire and our country a century from now,” he wrote.
The Oliver Wyman consulting firm has studied the impact driverless trucks might have on railroads and concludes the future likely is closer to Buffett’s vision than ARK’s. That’s the good news.
The bad news is that intermodal traffic would be highly susceptible to diversion once trucking costs fall below rail’s costs. Intermodal customers like J.B. Hunt, Hub Group, and UPS would put their loads on the road, killing the rail industry’s growth engine.
Wait, you say. How could highways handle all that intermodal traffic? Oliver Wyman contends a massive shift of intermodal traffic wouldn’t have much of an impact on rural sections of interstate highways. But there certainly would be more congestion on urban highways due to ongoing population and traffic growth. And intermodal traffic couldn’t escape this phenomenon, either, because of the need to dray containers from intermodal terminals to customer locations in urban areas.
Ultimately, railroads would lose their price- and service-sensitive intermodal traffic to driverless trucks. So railroads would have to fall back on carload and bulk business. Aside from chemicals, no carload segment has shown consistent growth over the past decade. But since carload and bulk business is highly profitable, it could sustain the industry for decades despite a lack of growth, Oliver Wyman argues.
Consultant Rod Case says railroads would not go away. Trains would likely become low-speed land barges that haul low-value commodities that aren’t particularly service sensitive. Railroads wouldn’t be dead, Case argues, just increasingly irrelevant.
I don’t gravitate toward doom and gloom outlooks, and tend to agree with Yogi Berra, who famously observed that “it’s tough to make predictions, especially about the future.” But change can move in unpredictable ways and come more quickly than expected.
A case in point would be photos of the New York City Easter parade taken on Fifth Avenue in 1900 and 1913. In the 1900 photo, you see lots of horse-drawn carriages and just one automobile. This is flipped in the 1913 image: You see lots of cars and only one horse.
Skeptics in 1900 probably cited all sorts of reasons that cars would never replace horses. They’re too expensive, they’re unproven and unreliable, and there aren’t enough gas stations or roads. We all know how that turned out. Yet parallel arguments are being made today about driverless trucks and battery electric technology.
The town where I live, Westfield, Mass., is nicknamed the Whip City because its factories were once leading suppliers of buggy whips. Of the 40 or so whip factories that once thrived here, only one remains in business. They were doomed to obsolescence once the internal combustion engine put horses out to pasture.
Let’s hope railroads don’t suffer the same fate.
You can reach Bill Stephens at bybillstephens@gmail.com and follow him on LinkedIn and Twitter @bybillstephens
He also notes that brick and mortar stores will go but people forget that humans have emotion and desire to engage….hence people will not be satisfied working and shopping from home. The decline of malls etc is mainly due to all the malls becoming generic bottom feeders with the same mass market stores. There was a time when malls had specialty and local stores. i.e. Lenox Mall in Atlanta had a fine Hobby Shop. Now it has Gap like every other mall has.
Rail already essentially carries many truckloads at a driverless equivalent. You have an engineer and conductor and combined the pay is within the same as 3 trucks with 3 drivers. Yet the train has 100+ cars so that’s close to 100 trucks in tonnage. Any gains a truck makes being driverless will be eclipsed by the same ratio if trains go crewless…considering the advantage trains currently have. I bet people thought the interstates and trucks would kill rail…certainly it must since trucks can turn around at any time and essentially drive anywhere on any road…essentially being a path…where rail is confined to a private engineered physical plant…with every inch having to be “heavily” engineered unlike side roads, driveways, etc. But instead of dying rail has grown significantly since the time interstates were started.
The word “autonomous” has been redefined surreptitiously. It now means that a vehicle (automobile, truck) can move itself, but always requires someone to sit in the driver’s seat to take over if (when?) something goes wrong.
So many reasons why the answer to this is a resounding no, and not just in the near-term, but for many years to come, if ever. Try this one on for size – do you really think the Teamsters are going to sit idly by while autonomous trucks put drivers out of business? That alone is good for at least another 20 years of delay – AFTER autonomous trucks are actually approved for prime time, which in itself is at least 20 years away.
My guess is that the biggest, er, roadblock to wide adoption of autonomous trucks will be not technological but legal and financial. Some mishaps will inevitably occur; and in this litigious society they will predictably lead to jury mega-judgments against the trucking company, the truck maker, the software provider, and, for all I know, the Icelandic Ministry of Fisheries. A few such cases might put the brakes on this technology.
WOW, you got that right. I call it the “free shot” . You can sue anyone and not win and not lose a dime in 99% of the cases.
One word answer to the question headlining this article, NO for all things mentioned above. But for those out there who think AI is the answer to all your needs then what is your purpose for being here?
I would remind everyone, especially Bill Stephens and the TRAINS editorial staff, that once upon a time railroads faced a similar technological challenge. The coal slurry pipeline was one of the biggest competitive threats in railroad history. But along came a “Visionary Railroader” by the name of Jervis Langdon, Jr. As B&O President he invented the coal unit train and put the pipelines out of business pretty much forever. Recommend my friend Roger Grant’s excellent book on Mr. Langdon’s life and career with the appropriate title of “Visionary Railroader” (Indiana University Press).
Mr. Wayman brings up a good point. Highway robbery increase in the future!!! As for cutting the cord. I have cable bringing in Wifi to my router. We pay only for the wifi. We use a digital antenna for local stations and get lots of tv channels through the wifi. We don’t buy any of the extras. Though we do on occasion pay for a movie through Vudu on our Friday family pizza and movie night. Not wasting my money on all that garbage.
Driverless truck or fleeting with one head-end driver has it’s limitations. Take the Grape Vine or Lookout Pass for instance. Trucks are typically down to 25 to 30 MPH on the twisting grades. The newer tractors with the Tier 4 engines are often on the side of the road with the hood up. What happens when the 3rd of a 15 truck fleet stalls going up a grade? Or worse yet, the driverless truck stalls in a canyon or tunnel and there is no satellite service?
Railroads and trucks are in the service industry. The railroads seem to have forgotten that simple fact. I mentioned this before, Claude Mongeau replaced EHH at CN. Claude mentioned in his interview that the first thing he had to do was work with all of CN’s customers on service issues. It’s the “last mile” that counts.
UP frequently runs monster freights down from Eastport, blocking almost all crossings in Spokane. However, the monsters are typically all unit potash or grain trains. Very seldom do the mixed freights exceed a miles or so, which is the typical length here.
BNSF typically only doubles coal and unit grain trains. Sometimes you will see a short block of auto racks or stacks tacked on to mixed freight. However, the cars capable of higher speeds, auto racks and Q/Z freights are not mixed with mixed freights. These higher priority freights provide more timely service to the customers.
What I have seen on UP, CSX and NS is everything under the sun on a monster freight. The challenge with that is the freight will only go as fast as the slowest car. So, if a mixed freight has a tank car carrying Dichlorophenyltrichlorosilane with a speed limit of 40 MPH, the high priority block of trailers tacked to avoid a crew start will only go 40 MPH. Sure, it will get to its destination 1 to 2 days later. However, if that is what your precision schedule called for, you were on time. Your customer, however, is still waiting for their shipment. Not good service. The customer will most likely pay a higher fee for a double driver and get their shipment in a third of the time.
With all of that said, driverless trucks and fleeting are not really feasible. Railroad service is pretty bad. And here we are.
There are thieves breaking into stopped containers in California. What is to stop them from getting a car in front of one of these unmanned trucks, slowing down and stopping. Then taking over it. Can it call 911?
Trucks vesus railroads? They all have their specific place and need. The right tool for the right job.
As an aside if the author of this piece “cut the cord”?
How is he getting his internet and how may streaming services is he subscribed to?
Tell us cost of the cable company bundle vs cost of a phone company and all the content providers.
Did he type this piece at home, put it in and envelope. affix a stamp and mail it to Kalmbach?
Does not going into a mall or local store just put money into Jeff Bazo’s pocket versus the communities?
I dont see any ATM’s dispensing coins or specific cash quantities for business needs as a brick or mortar presence.
Trucks versus railroad? They all have their specific place. Right tool for the right job.
This is another pipe dream just like the bizarre notion that the internal combustion engine can be replaced by having everyone driving electric cars by 2030 or 2035.
Both ideas ignore the hard reality that the infrastructure doesn’t exist to support the production of electricity to power all those new electric vehicles, or to construct millions of miles of new highway lanes to handle all the traffic.
Elon Musk pointed this out last year. Bad environmental policies are causing energy prices to skyrocket, our highways are already congested to the breaking point, and no new power plants are being built to supply all the electricity needed to fuel all the new electric vehicles. Wind and solar are too inefficient and unreliable to fill the ever-growing gap between supply and demand.
So go ahead, do the research, but don’t ask me for a dime of my taxes to waste on this nonsense.
Nowithstanding congestion issues, diverting intermodal traffic from trains to trucks would totally wreck the highway system. Trucks do not pay for the dammage they cause to roads and bridges, and are therefore massively subsidized by car drivers (through fuel taxes) and all levels of governments. Moreover, fuel taxes have not kept up with inflation, the highway trust fund has been insufficiently funded for more than a decase now, and its revenues will sink further as more road vehicles convert to batteries. I would not rely on sketchy, risky hedge fund managers to make accurate predictions about the future, be them ARK or any other. I’d rather stand with tried-and-true investing principles such as Buffet’s and Munger’s.
Totally agree with you on that one. When hedge fund managers get involved in anything it is usually the end of the line for that business.
“…fuel taxes have not kept up with inflation” More accurately, fuel taxes have not kept up with vastly improved fuel economy, which has drastically reduced overall fuel tax revenue. A car that gets 30 mpg (such as my Corolla, around town, not on the highway) pays half the fuel taxes of a 15 mpg pickup truck on a per mile driven basis. And as the anti-fossil fuel crowd keeps forcing the nation to adopt hybrids and all electric vehicles, that trend will continue. Not to mention those that will be forced onto public transportation because they can’t afford one of those hybrids or all-eletrics.
Teslas in self driving mode get into bad accidents, I cannot see how that can be safely implemented. Road conditions are too changeable.
My take is I am not comfortable with 80,000lb driverless trucks going down the highway in all kinds of weather and traffic conditions. In addition the highway infrastructure in this country is a mess even in good areas. Billions and billions will be required to bring the roads up to date. I cannot even fathom trucks in the congestion of Washington State especially in the SEA/PDX corridor or even close to metro areas.
So funny in Europe they run short fast freights on lines with passenger service, and trucks are often limited to right lane only with enforced limited speeds on limited access highways. In the US we operate super long trains at relatively slow speeds on lines with almost no passenger traffic and trucks on the freeways drive however they want.
All you have to do is look at Switzerland, not sure of the tunnel project but I seem to remember there was a plebiscite to pay a tax to build the railway tunnel(s) and force the trucks off the highways onto the rail. And short fast frequent trains.
The citizens did not want all the trucks on the highways.
It’s the service stupid! Transportation (which includes railroads) is a SERVICE business. If you want more customers, you give better service. And I would remind everyone that at Santa Fe we proved customers would pay more for better service, from the Super Chief to the Super Fleet. It is very interesting to note that the most optimistic rail industry leader/investor also owns and operates the ONLY railroad not to follow the lemmings over the cliff and endorse PSR. Good for Warren! He has a brain and knows how to use it.
Good point about BNSF.
They’re a control group placebo compared to the PSR drug of the other RRs.
Since BNSF is privately held, their books are closed BUT they’re pouring serious money into infrastructure, not the sign of a failing enterprise.
BNSF doesn’t use the Precision Scheduled Railroad term, but it practices PSR. You can watch 12 to 15-thousand foot trains repeatedly pass through Fort Madison, Iowa on Virtual Railfan. Yesterday on Broadcastify I heard a BNSF dispatcher refer to a 16,000 foot train being held for a relief crew at Galesburg. From the public grade crossing standpoint trains this long are not safe and should be outlawed.
Mark, one long train does NOT a PSR-junkie make. BNSF is smart enough to take the best attributes of PSR (and there a few) but leave the rest in the garbage can where they belong. As Warren told his shareholders this year, “You can be proud of your railroad.”
It sure seems to me that making railroads autonomous is a far safer and more effective step to take early in the development of this technology. It’s a restricted corridor (on rails), already has many of the pieces in place such as signaling and PTC and is less dependent on outside influences such as other drivers, obstacles and road conditions.
The “demonstrations” of self-driving trucks are currently on the scale of magic tricks, where something appears to happen under very specific circumstances. Just as that means you can’t just wave a wand and make money appear under a hat, it means that self-driving trucks are nothing like free and flexible vehicles. Look at all the Telsas that have killed people, both outside and inside the cars, when drivers assumed the cars could “drive themselves.” Having that range scaled up by several hundred times to a vehicle that could carry 80,000 pounds of hazardous cargo sound smart to anyone? For all the Hype, no free range self driver has hit S.A.E. 3 on the five point scale, which means not one of these toys is driving itself completely.
Better: self-driving trucks is really, really hard. Self-driving trains is surprisingly easy, and has already been demonstrated to work consistently, as with the light rail systems of places from Vancouver to Kuala Lumpur.
The minute someone decides that say a 75 – 80 foot Hi-Cube boxcar(or, lets think real big and come up with an articulated 100 ft Hi-Cube boxcar) filled to capacity of product delivered direct from manufacturer plant to warehouse sized direct to customer store(operated by said manufacturer thereby putting wholesaler out of business) and and entire world of physical commodities can move from OTR to rail at a much less expensive price point.
Gerald is a genius!!! Everyone else please pay close attention here. Several years ago, I worked on a consulting project (after leaving the Santa Fe) for a large appliance maker based in Benton Harbo, MI. They have moved most of their large appliance production to Mexico and were having enormous trouble getting trucking capacity to came north back in the US. We found the best railcar for them was actually an 85-foot auto parts car but most of these were too old and out of service. Also, there are some serious wear problems with the big design, so we ended up using 50 ft high cube boxcars. Not an ideal choice. Best car we could find would have been a dry box version of the jumbo refrigerated cars (used for frozen French fries) which are about 72 feet in length. Take out the insulation and refrigeration gear and it is a perfect car for large appliances, most of which are now produced in Mexico regardless of corporate brand. You need to ask KCS why they did not take our recommendation to buy the cars we designed and recommended.
James the answer to your question is KCS wanted YOU to purchase the car you wanted.
“…a large appliance maker based in Benton Harbo, MI. ” a.k.a Whirlpool yes?
For starters, ARK is hardly a dispassionate information source… they are part and parcel with the whole Silicon Valley nonsense. They are SELLING something, and some of it is vaporware hype.
Level 5 autonomous vehicles do not exist and it is doubtful they ever will. Level 4 might be feasible, but this likely requires geo-fencing. And in place like Central NYS where I live, I am highly skeptical with the poor physical condition of many roads (including interstates) and variability in weather are extreme technical challenges.
TuSimple is running tests on I-10 from the outskirts of Tucson to the outskirts of Phoenix — I used to make this drive many times per month. The road is basically flat, has few curves and weather is seldom a problem. This is like Waymo’s geo-fenced taxis in Chandler, AZ — hardly “the real world”. Remember, of course, all the hype of full level 5 autos from Tesla, Waymo and Uber? Still waiting, 5 years after the promised date.
I didn’t see much about what operating speed TuSimple uses or is promising. From what I’ve read, the optimum speed for level 4 trucks is 45 MPH, which is hardly compatible with interstate highways.
Then there is the small matter of insurance. Who is going to assume the risk associated with killing people with these things?
The only real threat to railroads is their adoption of asset-stripping PSR. This is short-termism on an epic scale. Facilities have been ripped out, often as symbolic acts of madmen like Harrison, workers tossed out in the pandemic or PSR who have no interest in returning to draconian working conditions, and a virtually complete failure to invest in meaningful capacity additions (the CSX Gulf Coast drama best exemplifies this phenomena by running 15,000 foot trains that cannot use any siding).
Agreed. Electrify the rails overhead, like the fast freights in Europe and cut the PSR trains in half with more scheduled frequency, then I believe there won’t be any worries for the next 150 years. Highways just can’t beat the efficiency and built-in flexibility of railroads.
When the first family on a road trip vacation is murdered by a driverless truck public outrage will end the argument. Cars sharing the road with heavy driverless trucks, what could go wrong?
Right. Those families should be murdered in traffic the way God intended, by people driving poorly, or under the influence, or an with unsafe vehicle, or speeding.
I have no idea of what the future holds, but I do remember when 747s first rolled out there were doom and gloom predictions about them taking automobile traffic from the railroads, so…
The energy equation and autonomous unmanned freight trains, which are well along in the development phase, will keep the railroads competitive. Single-person crews are simply a step along the way. The current roadway infrastructure cannot handle more tonnage without an additional prohibitive investment of public funds, even with President Biden’s current infrastructure investment legislation. The railroads as private companies have to pay most of their own infrastructure improvements up front as they own the land the lines are on. That allows governments to limit direct handouts of public money to the railroads. As usual the railroads will go begging for tax breaks to maximize an already rich bottom line if that happens. The 4-to-1 energy cost advantage is huge. Eliminate manning and human inspection costs and automate the equipment inspection processes using artificial intelligence-based wayside inspection , which railroads are already attempting, and you get into the sub-45% operating ratio range. The truckers cannot match that even with a huge influx of public money. The truckers will be competitive inside 1,000 miles because the railroads have abandoned that traffic. Transcontinental movements are a different ball game. Mr. Buffet and his railroad are OK for another half-century, if he lasts that long!
With all that additional tonnage on rural interstates crowding out passenger vehicles and leading to quicker road deterioration than is currently taking place who will pick up the costs associated with this diversion of freight traffic? And how do driverless trucks respond to blizzards and other weather factors? Safety could be a major issue unless trucks had their own toll road systems that were closed to passenger vehicles.
Have to agree with you Mr Gishlick. I don’t know where you are located an if you saw the the pile up on interstate 81 in PA. yesterday March 28. There was a snow squall around 10:30 am. Over 60 cars and trucks involved three dead and over 20 hurt. Video showed vehicles running into the pile up at highway speed. Myself I don’t like the idea of a 80’000 lbs truck behind me with no one in the cab.
Perhaps, but a bunch of “trained and licensed” drivers of a number of vehicle types also screwed up. An AV might just have enough brains to recognize weather conditions and slow down.
Railroads may benefit from self-driving trucks at intermodal terminals adjacent to ports providing near dock rail. Beyond that, public policy may weigh in to restrict these driverless trucks.
As I continued to ponder this, I was stunned when I considered that of all the various modes of transportation, only railroads run at a lower average speed than they did 100 years ago. The replacement of the steam engine that loved to run at high speed with the diesel electric that excels at lower speed was actually a step backward. What new money saving form of motive power will railroads come up with to go even slower? They have already got PSR, an inflexible operating mode focused on only the operating ratio.
For starters, DPU’s on every train… for the PSR hive-mind, just an excuse to run impossibly long, poorly-built trains that have no hope at maintaining track speed.
If you’re measuring success by how many trains you put engines in the middle of, your priorities are a liiiiitle bit misguided.
Railroads will be as irrelevant as they want to be. This means they will likely make themselves the carrier of last resort for heavy bulk items that make no sense to move by truck.
The whole approach of railroading is big power, big cars and long trains. Fast and quick is an anathema to railroads. Yet quick, fast and reliable is what many shippers are willing to pay for.
Railroads and railroaders have always lived in a world apart and tried to make the world conform to them. It worked in the beginning, but it is working less and less as time goes on.