Modeling self-driving trucks: A 2025 report and my predictions from 2017
(This post seems long, but only because I copied three 2017 posts at the end.)
I came across a great article, “The Outsize Impact of AI Logistics: Automated logistics will heav[ily] impact one third of GDP” by Austin Vernon, published on July 5, 2025.
That article inspired me to dig up the series of articles I wrote in 2017 about self-driving trucks. I wanted to see if my predictions held up!
In 2017, self-driving trucks seemed right around the corner. Although there was some discussion about how they would impact the supply chain, most of the discussion was about how many truckers would lose jobs. When I wrote the articles, I wanted to provide a different viewpoint and emphasize how it would impact supply chain modeling.
Before I get to my 2017 articles, let’s talk about Vernon’s article.
Self-driving trucks seem like they could be here within five years— I’ve personally taken a Waymo ride and know people who use Tesla’s self-driving. But, in 2017, we might have incorrectly predicted self-driving trucks within five years!
Vernon discusses more than just self-driving. He uses the term “AI Logistics” to mean self-driving vehicles, drones, and robots. He also adds electric vehicles to discuss cost advantages further. This shows that technology has advanced even since 2017.
I’m going to figure out how to use his article in my supply chain class. He starts with the importance of logistics: It impacts 10% of the GDP directly and heavily influences another ~25%. He does a great job of breaking down the current cost structures of various modes of transportation and how self-driving electric trucks will drastically reduce those costs.
The article has interesting discussions of how this will impact logistics: a variety of trucks for different-sized loads, more point-to-point routes, a collapse in LTL costs, and a possible impact on where you locate factories.
Luckily, his article fits nicely with my 2017 articles. You can see the full articles (with the links) at the end of the post. Below are the points I made in each article. You can see if you agree that they still apply.
In “A Non Obvious Way Self Driving Trucks May Impact Your Network Design Strategy,” I argued that cheaper self-driving trucks may mean that you spend more on transportation to improve service and reduce inventory. For example, it is less painful to carry inventory at many locations because, with small transportation batches, it is easy to replenish or transfer daily.
In “Why Driverless Trucks May Create the Need for More Drivers,” I applied the ATM (see below) example to trucking. Maybe the cheap self-driving trucks for long-haul will create opportunities for more high-touch direct-to-customer drivers.
In “Self Driving Trucks May Create New Roles and New Types of Jobs,” I point out the other jobs that will be created around self-driving trucks— like monitoring, maybe more brokers, and, now that I think about it, more optimization and matching engines.
I’ll end with a new prediction: As self-driving electric trucks and AI logistics robots in warehouses and on loading docks start to come online, we’ll need to redesign our supply chains with new and better supply chain models.
For students reading, this sounds like a good research project. For practitioners, keep your modeling skills sharp.
Original 2017 Articles Below
Originally published on January 10, 2017, at Supply Chain Digest.
A Non Obvious Way Self Driving Trucks May Impact Your Network Design Strategy
The Driverless Truck Phenomenon is Closer to Reality and That Means it’s Time to Start Strategizing how This Could Impact Supply Chains
Driverless trucks could have a large impact on the design of your supply chain. Jeff Potts of LeanLogistics wrote a nice blog post that hit on some of these ideas. He gave us a few good insights:
Driverless trucks open up many new possibilities for your supply chain and the ways you can serve your customers and compete in the market.
Driverless trucks promise to reduce transportation costs by avoiding getting more use from each truck without having to worry about drivers needing rest.
Driverless trucks will likely first start out on the interstates and may not do the first or last mile of the trips.
The old rule of 400-500 miles per day no longer applies. He has some nice maps showing that 2 warehouse can now cover a lot of the US within 1-day versus 5 warehouses today.
Will the ability to have a bigger one day service territory reduce the number of warehouses? Maybe.
But in optimization, the trade-offs are more subtle. Driverless trucks reduce the cost of transportation. If this cost is lower, you may want to spend even more to reduce the now higher relative cost of long service times or inventory by an even greater amount. Let me explain.
First, I’m assuming that driverless trucks will be cheaper and easier to operate within your own network. That is, the easiest thing to try first is to send a driverless truck from your own plant or central warehouses to one of your own local warehouses. That is, initially, it might be better to have a driver show up at your customer location to deal with their technology and all the rules they may impose.
Second, I’m assuming that driverless trucks will quickly be built in different sizes. With a person driving, you had an implied fixed cost so you wanted to maximize your use of trucks that held 40,000 pounds (or your equivalent in cube or pallets). But, what is to stop the proliferation of trucks that are built to carry smaller loads (and use less fuel)?
Given all of this, it may make sense for your network to have even more warehouses. The extra warehouses can improve your service even more-- think about being able to do mostly same day delivery from your local warehouses.
With cheaper transportation costs (both from not having a driver and by being able to efficiently ship in a truck built for your load size), it is easier to replenish many warehouses. Also, the cost of inventory usually prevented having too many local warehouses. But, with much cheaper and faster replenishment to these warehouses, you can replenish these warehouses daily from the central warehouses. This keeps inventory to a minimum and allows to you provide even faster service to your customer with much later cut-offs.
In this case, maybe you spend 10% more on transportation and warehousing costs, but you reduce your lead-time to your customers by 80% while having minimal impact on inventory levels.
And, if you are currently sending product to your customer’s warehouse, this local inventory may allow you to deliver directly to the stores, job sites, or other customer locations.
Final Thoughts
Driverless trucks open up many new possibilities for your supply chain and the ways you can serve your customers and compete in the market.
Originally published on January 24, 2017 at Supply Chain Digest
Why Driverless Trucks May Create the Need for More Drivers
Less Expensive Transportation may Lead to More Local Warehouses That Offer Same day Delivery and More Direct Customer Delivery
There has been a lot of talk about how driverless trucks may reduce the need for drivers and how will our economy replace these 2 million good blue collar jobs.
The same conversation is going on for the larger economy with all the buzz around artificial intelligence.
Lots of smart people and deep thinkers are worrying over the issue, and I don’t have a strong opinion in the debate.
My guess is that the changing economics will lead to whole new ways to think about your distribution network.
However, I heard an interview with someone recently (You can find the interview here at the 15:44 mark) who compared today’s fears with the fears people had when banks introduced ATMs. At that time, there was a fear that ATMs would replace all tellers. What ending up happening is that there are now more tellers. ATMs changed the economies of scale. Previously, banks had to be large enough to handle the customers. But, with ATMs, you could build more and smaller branches. So, for sure the tellers per location went down, but the total number of locations went up enough to increase the total number of tellers. It should be pointed out that the tellers also had to do a slightly different job. And, the customer experience was now much better-- it is likely you are now much closer to your bank.
Could all this happen with trucks?
In a previous article, we talked about how driverless trucks could easily change the design of the supply chain. With less expensive transportation, you may have many local warehouses and offer more same day delivery and more delivery directly to the customer.
That is, instead of paying drivers to make the long haul from my warehouse in Chicago to my customer’s warehouse in Dallas, I let the driverless truck make that long-haul move to my new local warehouse in Dallas. From there, I send out a driver to make the local round of deliveries and maybe even directly to my customer’s stores.
In this case, there will be far fewer drivers doing the multi-day long-hauls across the country. But, I may have many more drivers making local deliveries on short notice. Like the bank tellers, the driver’s job may change. Instead of just getting a load from point A to point B, the driver may act more as a face to the company and participate more in sales and customer relations.
This assumes that the last mile is going to be hard to automate -for example if you are Coke and Pepsi, it seems like a long time before a driverless truck shows up at the local supermarket and figure out how to get product onto the right shelves in the store, while picking up competitive information.
Final Thoughts
It should be interesting to see how all of this shakes out. But, my guess is that the changing economics will lead to whole new ways to think about your distribution network. And, some of these changes may mean that you use more drivers.
Originally published on February 7, 2017, at Supply Chain Digest
Self Driving Trucks May Create New Roles and New Types of Jobs
What Will the new Role of Drivers be When Self-Driving Trucks Start Hitting the Road?
The above picture is one I took while on vacation in Costa Rica. We were going to a remote beach at the end of a 3-mile pothole and rut-stricken dirt road. We happened to be following a truck making a delivery. The truck got stuck by an overhanging tree. The driver immediately got out, with a large machete in hand, hopped on the top of the truck and proceeded to cut away the offending branch. Within a few minutes, he was back on the road.
I was thinking that it was going to be a long time before self-driving trucks have built in machetes on the top. And, some people may even worry about such a feature. I have the feeling that this job on this route is safe for quite some time.
Most places trucks go don’t have a need for the driver to clear a path with his machete. So, what will the new role of drivers be when self-driving trucks start hitting the road?
In a previous post, we mentioned that self driving trucks may actually create more jobs for drivers-- as companies rely more on drivers to make the final delivery, help with sales, and gather competitive information.
Another new role we anticipate is the need for people to remotely monitor each of the trucks on the road. I picture a big war-room with 100s of monitors showing what is in front of each truck, what is behind it, and maybe what the load looks like. The controls will give the operator the ability to have two-way communication to talk to someone near the truck, the ability to provide new direction, and ability to see and respond to alerts or problems. This will be needed to prevent theft of cargo and to allow interactions at key points. For example, you may need to talk to someone at a re-fueling station, deal with an policeman who pulls over your truck, or even negotiating to find a spot to park the truck when the warehouse can’t unload it. A single person may be watching over ten to fifty of these trucks at time. This obviously won’t replace all the jobs of people in the trucks, but will replace some.
There may also be a bigger need for brokers to match empty trucks with with loads. Truck owners that keep the truck making productive moves 24x7 (with just stops for re-fueling and maintenance) will be much more profitable. To do this, you will need brokers to help find these productive moves. And, since the cost to the shipper will be lower, shippers are likely to want to move even more loads-- they won’t wait to fill up a truck. And, there are likely to be more opportunities for matching trucks to loads. For example, with drivers you have to worry about running out of time. But, if half full truck is passing by a load that would fill up the other half, it is much easier to pull over and get that load. You may add a few hours to the delivery time, but you don’t have to worry about having to rest for another 8-10 hours.
Final Thoughts
I’m optimistic about the future of self-driving trucks. I don’t think the 2 million truck driver jobs will just disappear. I think the nature of these jobs may change and it could be painful for the drivers who are directly impacted if they can’t move into new roles. But, I think self-driving trucks will open up a whole new set of opportunities.