On Trucking
snowingpine.com
snowingpine.com
He rode across the country with an owner operator, and he tells the story.
"A true truck has eighteen wheels, or more. From Atlanta and Charlotte to North Powder, Oregon, this was the first time that Ainsworth had so much as tapped his air horn. In three thousand one hundred and ninety miles I rode with him he used it four times. He gave it a light, muted blast to thank a woman in a four-wheeler who helped us make a turn in urban traffic close to our destination, and he used it twice in the Yakima Valley, flirting with a woman who was wearing a bikini. She passed us on I-82, and must have pulled over somewhere, because she passed us again on I-90. She waved both times the horn erupted. She was riding in a convertible and her top was down."
I love the ambiguity of the last sentence you quoted. I won't spoil it for others; just go read the article, it's that good.
Fair warning: it is long form journalism, of the best kind. Give yourself some time to savor it.
Now that is a great ambiguous sentence.
https://www.newyorker.com/magazine/2003/02/17/a-fleet-of-one
My company, Platform Science, is building a configurable open platform that connects a lot of the different categories mentioned in the article (drivers, carriers, brokers, OEMs, software vendors, etc). If you're interested in career opportunities in the trucking industry, we're hiring for a number of open positions[1]. Feel free to send me an email (My work email is in my profile) and I can connect you with the right hiring manager.
Platform Science looks pretty cool - seems like a TMS with specific improvements against legacy systems. Building a TMS is a pretty good business - massive fixed cost but once you've built the system, it's fairly hard for competitors to dislodge you.
A couple of questions would be:
* At what speed is the industry going to embrace self-driving / "augmented driving" in trucks? I'd imagine a driver would still be required at the wheel in case of emergencies for the foreseeable future. What new skills would truckers have to learn?
* Given the environmental impact of trucking (self-driving or not), what are some interesting propositions that climate change startups could come up with? Do any players in the value chain even care about climate impact? Clearly, given the cut-throat nature of the business, there would necessarily need to be a revenue impact otherwise no one is going to use the proposition.
* I find the insurance part more interesting that the factoring part. Given the generally poor service quality of insurance companies, I'm wondering if there is a possibility of a "Lemonade" like insurance company in this space? Lemonade started out as a home insurer I believe. What is the killer proposition for an insurance company in this space?
Lots more thoughts come to mind. Again, great writeup, thanks!
Yes, and manual work for on/off-loading or even paper work is still not uncommon. I can imagine that self-driving on long highways will come as a first step to avoid longer stops (no sleep time required) or to avoid a second driver.
> Clearly, given the cut-throat nature of the business, there would necessarily need to be a revenue impact otherwise no one is going to use the proposition.
I was wondering why route optimization wasn't mentioned in the post. Lot's of innovation happened in the past years. (With "route optimization" I mean both, the route selection between two points and also optimization of the order of multiple stops.)
ICE trucks are really hard to operate. You are basically micro managing the gears, dealing with lot of fairly complex bit of equipment, etc. Electric trucks basically have no gears, heaps of torque and don't really require much more than just pointing them in the right direction. Apparently as easy as driving a normal car.
So, electrification strikes me as a core requirement for self driving. And of course electric trucks are also more environmentally friendly. Less brake dust (regenerative braking) and less emissions. And the batteries are recycled at the end of their life. So there's that.
The shift to electric trucks is already happening of course but its very capital intensive. Millions of trucks with a price tag of hundreds of thousands each means a lot of capital will have to be expended. Trillions basically. That alone will ensure it will take quite long to complete. A decade minimum before most new trucks are electric and then at least another decade before most ICE trucks have been retired. And the first generations won't be self driving. So these new trucks would be in service for a decade plus before being swapped out for self driving ones. So, if that takes another decade to happen, we're talking 30-40 years. Minimum. But with the first ones on the road as early as less than a decade from now. It's going to be a slow process.
Insurance is basically about risk. Insurance companies like to minimize risk and maximize profits. An electric truck that is less likely to break down regularly is less risky and would on paper require a lower insurance fee. And then a self driving truck is less likely to get into accidents or other liability related issues. Which again means lower fees. So, that helps.
Volvo (which is BTW one of the more optimistic companies) is still today excluding long-haul trucks from being fully electric, and they have just started production of short-haul and "regional" electric trucks.
And of course full self-driving is still beyond the horizon.
BTW, and only as a side note, traditional trucks already use engine braking or retarders extensively, so I don't think that regenerative braking will change much when it comes to the amount of brake dust.
Hybrids seem to me like a net win all around for at least some use cases. I'm not sure where the value proposition is breaking down (or if anyone's even put in the effort to figure out whether it actually is breaking down).
Surely there can be particular routes/scenarios where hybrids may offer an advantage, but it is not (at least to me) very clear which ones they are.
Using a hybrid, you get back a lot of the maintenance hassle and on top of that your fuel is again more expensive. It's basically most of the downsides of an ICE truck without most of the upsides of a full battery electric truck. So, the business case for hybrid trucks is just a lot more murky. I'm sure there will be some companies that will try this but the money is going to be in full battery electric.
> you get back a lot of the maintenance hassle and on top of that your fuel is again more expensive
The entire value proposition of hybrids is that they both reduce the amount of required maintenance on the ICE part (plus the brakes) and reduce the amount of fuel needed-- while maintaining all the flexibility of ICE vehicles. As a bonus, they require far fewer batteries than full electrics at a time when battery supply is clearly going to be constrained for the foreseeable future.
Fully electric trucks will have trouble anywhere with long stretches of cold weather. Operating outside of their ideal temperature zone will result in much lower efficiency, exacerbated by the weight of the batteries remaining the same regardless of charge level.
I think there are ideal niches for both approaches. Honestly, I also think most pure ICE drivetrains should be converted over to varying types of hybrids in the long run, because it almost always seems to be a net win to me... but I could be wrong. Would love to see more R&D put into it, in any case.
Is that a net win? It reduces maintenance on the brakes, it may reduce maintenance on the ICE, and it adds a whole bunch of new stuff to maintain. Does anyone have data on the net effect? Does anyone have data for trucks?
Nikola advertises a battery electric heavy truck with 330 miles of range. Nikola and Volvo seems to still be pretending that they are going to use hydrogen but at the same time they are selling a lot of battery electric at this point and not a whole lot of hydrogen trucks. Mostly ranges are creeping up to the point where drivers must stop to take a break anyway that is long enough to top the battery up. So, the range anxiety argument is simply not there. It's a pure cost driven thing. And it's not like hydrogen fueling is particularly fast or cheap.
Of course, a lot of companies are opting for shorter range trucks because they are much cheaper and because that suits their needs. Long haul electric trucks are more expensive. Because of the extra batteries. So, yes, the sweet spot of the market is short haul, for now. But there's nothing inherently hard about longer range trucks. It's just a cost and supply chain issue. As battery production scales and prices come down further, we'll see more companies dip their toes into long haul.
Self driving is a thing that multiple electric truck and van companies are talking about at this point as something that they want to do. It's still pretty far out but it's not that much harder than self driving cars. But as I said, the bigger challenge here is simply replacing fleets which is an inherently slow process. But of course that's exactly where the money is going to be for manufacturers.
Not entirely unlike existing multi-trailer (used on some roads in Australia), so called Road Trains, but limited to two trailers, the main one and the battery trailer:
https://en.wikipedia.org/wiki/Road_train
the truck would enter the highway, stop at a service station, add a "powerpack" battery trailer, drive until the next station, swap battery trailer as many times as needed, then leave it at the last service station before destination.
It will fail in the market because the technology does not exist and will not exist in the foreseeable future unless or until something like AGI is perfected.
Driving is not a closed domain problem. There will always be fat tail circumstances where statistical methods will break down.
Furthermore driving, freight haul and taxicabs are NOT natural monopolies. They are not areas where technology can give enduring market power that society will tolerate (except perhaps in the case of regulatory / institutional capture). Instead they are low margin, established, service- commodity businesses.
The point of trucking is to safely and efficiently get goods from A to B. Until machines / automation can do it cheaper than humans it will not be viable.
There is no shortage of cheap low skilled labor in the USA.
Why invest in hugely expensive automation in an open problem domain that you can not even perfectly model when you can just pay a driver $25 an hour to do it better?
There is so much more easily attainable low hanging hanging fruit.
Why not automate warehouses?
Trains? Waitstaff? Insurance agents? Real estate agents?
Triage nursing?
All these are probably more attainable.
The driverless delusion reeks of sophomoric understanding of the problem domain fueled by credulous media and group think.
Personally I wouldn't hire anybody for a thinking role who believes that driverless trucks or taxis are viable marketable product for the open road. It is a good heuristic to separate critical thinkers from bandwagon riders.
It is literally an easier technical problem to land a man on the moon than to design an autonomous road vehicle that can do better and be more cost effective than humans in all circumstances.
Hopefully we will stop talking about this foolishness in a couple years.
https://blog.piekniewski.info/2021/12/18/farcical-self-delus...
Some of it is. Walmart has already deployed some self-driving semis in Arkansas that go between a warehouse and a store. There are a significant number of routes like this, always taking the same path back and forth between two places in a relatively controlled environment.
> Why invest in hugely expensive automation in an open problem domain that you can not even perfectly model when you can just pay a driver $25 an hour to do it better?
Driving time is a big one. There are strict limits on how many hours drivers can be driving on a given day. That means you've got a huge, expensive machine sitting idle for most of the day. Self-driving can get you double the usage of an expensive asset.
> Why not automate warehouses?
This is happening already. https://www.wsj.com/story/amazon-takes-steps-toward-warehous...
> It is literally an easier technical problem to land a man on the moon than to design an autonomous road vehicle that can do better and be more cost effective than humans in all circumstance.
Nobody's saying it has to work in all circumstances. You can choose where to deploy self-driving vehicles. If you stick to the American Southwest, you still have an enormous opportunity, but you eliminate one of the big challenges for autonomous vehicles - snow and other difficult weather.
For example, fully autonomous trucks which only work on highway with special equipment on the side could already be worth it. Long haul drivers can’t drive without stopping, can make mistakes and it’s not a very pleasant job. If you could only use a driver for the last few miles, that’s a huge gain.
You are mistaking a failure of your imagination for an absence of possibilities.
It doesn’t have to solve all cases to be immensely useful. All it needs is to be able to recognize when it’s unable to safely have control with enough advance notice to delegate to a human. Remotely piloted aircraft have been a thing for a long time; seems like we could do the same with trucks, so that humans only need to do the driving a small % of the time. I also suspect that current AI will likely respond to emergency situations better than humans do on average, because they don’t get tired and have superior reaction time.
I'll alter one of my favorite jokes to explain.
I want to go out like my Tesla ai did, in a bit of confusion and reanimated from rom, not screaming no no no no like the people in the car.
Tho truly with a lot of these the people will go out in their sleep.
IMO, there are road conditions where this is possible. For black swan cases like random wind gusts, tire blowouts, boulders falling onto the highway etc. I think AI would be better on average *currently* (again, just clarifying what I already said).
The most likely beachhead will be desert lanes in places with low regulations where the middle mile is autonomous but things are handed off to a real driver for getting to the actual warehouse. That’s complicated/limited use cases/more expensive for the same product. Tough sell.
I work directly in heavy duty trucking tech. Our customers are the biggest fleets in the country. All of them are exploring the tech to not fall behind, but behind closed doors every trucking leader will admit we’re nowhere near operationalizing self driving. Not to mention it’s unclear if the biz model for autonomous trucks is to sell the tech (then who is liable when the robot kills someone?) or to sell trucking capacity, the latter squarely making the autonomy companies competitors of trucking carriers vs vendors. It’s messy as hell.
Self driving is the VR of logistics. “We’re a year away” for the next 25 years is where my bets are being placed.
2 - Unfortunately, the margins are too thin for owner operators to have a vested interest in climate. I do think that an interesting distribution strategy for cleaner trucks is to sell heavily into private fleets (e.g. walmart trucking) because there's an ESG angle that the companies can buy into. As for dedicated trucking carriers, it's going to be hard to convince them unless there's some realistic return on investment.
3 - This is interesting, I haven't looked into insurance too much personally and I know the barrier to entry there is extremely high. That being said, I think the liability aspect reduces the viability of offering low-cost insurance. Most trucks are essentially houses on wheels in terms of cost and a single accident can run in the millions.
I'd imagine that if you really wanted to build an insurance company for trucking, you'd need to focus exclusively on mega-fleets (25+ truck carriers) and somehow offer them better rates than legacy insurance companies. While owner operators pay a lot in terms of insurance, the market simply isn't big enough to build a cohesive risk engine and underprice existing insurance without taking on unaccounted risk.
Solar panels are (new) around $0.75 / Watt and less than that used. Maybe you could fit 9 or so on the top which, depending on what panels you use, could be about a 1kW or so in sunlight. Maybe it's too much of a stretch to ask for a modified engine but heating the driver cabin or even providing some type of heating or cooling to the container might help offset costs.
[0] https://www.ttnews.com/articles/nacfe-issues-confidence-repo...
My second thought is that even if the math worked out the incentives might not since it's a combination vehicle and each part of the combination is probably owned by a different entity. That's where a government mandate might be needed.
It doesnt have to be moving to charge. Provided its parked outside ofcourse.
But whether it pays to have to move around the extra weight for the energy produced is an important point.
That being said, solar panels are great but what happens when they break down or get damaged. I don't think the savings are worth the hassle right now.
The value is lower than putting the solar panels in a desert where they don't shake around and get worn out quicker. Perpetual solar powered auto-mobiles weight almost nothing and are more like three wheel bikes where you lay down. So on a truck it would be bugger all.
Here is some info on the Jones Act for those who are unfamiliar: https://www.cato.org/publications/policy-analysis/jones-act-...
Would be really curious for insight into why truck purchase loans run 8-10% APR with 20% down. If anything, I'd expect them to be lower risk and easier to re-posses?
(family member is a truck driver)
Also, repossessing trucks across the country is kinda an issue. You might need to drive hundreds or thousands of miles to claim the truck which is a nightmare that carriers deal with on a daily basis.
It can be difficult to locate a tractor when the operator who is in default on the loan/lease has all of the lower 48 to make it disappear.
Good analysis of the nooks and crannies of potential for trucking. I work in this space and there's still a lot to learn here.
Perhaps looking at trucking from a human centered design perspective would identify problems and needs which are missing from this otherwise deep analysis. What are the pain points of the various types of roles/people involved, what makeshift solutions are they currently using, and do they have money to spend on a better solution?
That being said, human centered design probably doesn't work in trucking. The assumption here is that people within a certain role has a fixed set of pain points. The truth is that trucking is too wide to categorize in that way. E.g. I met dispatchers that had a really hard time finding loads, but others had a really hard time getting drivers to pay them. Every person's experience and pain points in trucking are all slightly different simply due to the nature of the industry.
I think there's certainly an education component to most trucking companies failing but the root of the problem is the design of the industry.
Am I correct in naively calculating something like $200B/year on the table for full-self-driving truck technology?
Sure I guess there are operating models where you don't need one driver paired all the time with one truck but definitely can't just replace drivers IMHO
There's no reason, but then it also proportionally reduces the cap of $65k to some lower amount, that doesn't include those extra responsibilities.
Might need to add the capex of equipment costs, and subtract the externalities costs of increases wear and tear on infrastructure and equipment.
Also, some of the easy use cases for potential self-driving tech have - at least in the US - already been put on rail which is in a sense almost a self-driving truck. In my view one of the major opportunities missing for making trucking more cost competitive and reducing the societal impact of trucking (emissions, road wear, less than desirable work conditions for truck drivers) is to move more and more freight onto rail. The US is quite good at it actually, but in particular Europe needs to get their act together on it.
How did the spot market work before the internet/the proliferation of technological devices? Where did the "pool" of owner operators hang out before the internet? Or was trucking completely unlike it is today?
I'm having a hard time picturing either of these methods. It feels hard to have a large spot market of owner operators without some large broker in the middle, at least pre-cellphones.
Sounds a lot more human too. Interesting!
So I guess these drivers mostly hung out at truck stops (they still do today), just that most truck stops aren't as clean/nice as before.
Maybe the disruption this market needs is a set of unions rather than just trucking companies.
It would be nice if unions were limited such that... E.G. they could not grow to larger than 15% of the workforce for a sector in any single union. Contracts would need to be approved by... some number, lets say 2/3rds of all workers across all unions. With the 15% number that would require at least 5 unions and over 66% of workers to approve of contracts.
[1] Sincere praise, not sarcasm.
After reading that, you can safely ignore any comments for this person regarding European unions. I can guarantee you it’s an American and they have no idea of what they are talking about.
Unions in Europe have little to do with their American counterparts. Membership is never compulsory to begin with. They are just representative of labour for negotiations. Part of their role is actually mandated by the state. For example, minimum salaries for each roles here is set by bilateral agreements between the biggest employers and employees unions.
They play their role fairly well and nearly never strike. Calling them a cartel is beyond ridiculous.
> representative of labour for negotiations
in your words they try to get as much of the economic surplus for themselves as possible. They can do this with a long time horizon like German unions where there’s s consideration that you want the company to still be around in 30 years so your child has the option to work there or American style give me all your money right now short termism.
Cartel isn’t a term of abuse. It’s a description of the entire point of a union. They monopolise labour and negotiate with employers to the benefit of the average worker. The less like the average worker you are the less well they’re going to represent you.
Don’t worry I studied economics too.
You are confusing value added and economic surplus. That doesn’t bode well for the rest of your argument.
A change in labour costs is a shift in the equilibrium point. It doesn’t come from a surplus.
> Cartel isn’t a term of abuse. It’s a description of the entire point of a union. They monopolise labour and negotiate with employers to the benefit of the average worker.
Of course it is a term of abuse.
Collective representation isn’t about monopolising labour as a casual glance to any union in a western country like France would have told you. In France no union has more than a measly 2% of the work force. Yet they play an important role.
Or at least have tracking software that fines them heavily for leaving the slow lanem
You’re either driving too fast or too slow. You try to stay in the slow lane but catch up to the person going 45 on the interstate who speeds up as soon as you get next to them trapping you in the fast lane until you give them ‘incentive’ to let you back over.
I could go on and on but my government mandated 30 break was over six minutes ago and there’s people to annoy on the two lane highway I get to drive on until Texas.
I’ve also have seen some truck crashes where I’m like “how is that even possible?”
The thing with truck drivers is the bad ones don’t last very long. It’s almost impossible to get another driving job after getting fired for being a shitty driver.
—edit—
And law enforcement holds professional drivers to a higher standard so there’s no talking your way out of a ticket if they see you doing something stupid. Most of the time that’s when they get all up into the logbook and full on level 1 DOT inspection on the truck.
So, for more context on factoring, there are two types - recourse and non-recourse. Recourse factoring means that the factoring company can come back and ask the driver for their money back. Generally, the rates for recourse factoring are lower (by 20-50 basis points only because the default rate in drivers is quite high).
Non-recourse factoring is something that the industry is transitioning to. Most of the startups offering factoring today are doing non-recourse. The problem with that approach is that fraud rates rise exponentially. Not too sure how the landscape will look even 6 months from now as easy capital goes away.
* https://podcasts.apple.com/us/podcast/the-trucking-episode-w...
* https://omny.fm/shows/odd-lots/the-trucking-episode-why-the-...
Interesting observations on tracking in that episode:
> Gord: (23:44) […] So they regulate the drivers, right? And there's going to be another book being published next month by an academic at Cornell named Karen Levy, which is called “Data Driven: Truckers, Technology and the New Workplace Surveillance.” And she has spent 10 years studying like the effects of all of the regulatory imposition on the driver through surveillance technology, driver-facing cameras, ELDs, all of this stuff that's meant to regulate us because of the public's fear of, you know, drivers driving tired over their hours, all this stuff, which is an effect of the fact that the market's been pummeled by deregulation. So there's this regulation question, but it's not looked at correctly, that they've deregulated the market, but they just moved the regulation from the operations of the companies and the rating and the business side of it, and they've moved all the regulation onto the operations and onto the driver, which is another reason people wash out.
> Because if you're somebody like me who's been in the business my entire life, my dad was a trucker, both my uncles were truckers, my grandpa was a trucker. I was helping mechanics fix trucks and driving around when I was a teenager after school. And I'm one of the sort of last of the big game hunters. I know what I'm doing. If somebody is a professional and knows what they're doing, they don't want to be told how to do their job by some human resources harridan, or a health and safety pencil neck person that's breathing down their neck. And that's a factor.
> Another thing, and this applies to more than just trucking, is like the psychology of people who work for a living. Most people that work for a living just want to do their jobs and be left alone. And we have this management mentality where like every single thing has to be done exactly as the computer models tell us and as safe as possible. And they're trying to impose theory on material reality and it drives the people actually doing the work insane. So, you know, you want to end driver churn and driver retention? One of the factors causing that is that they've overregulated the people doing the work rather than the people in charge of the markets in which the work is being done. Does that make any sense?
> Joe: (27:20) Yeah, and that's an incredible point, and I hadn't really thought about this before, this idea that it’s like, okay, the business of trucking, the business of pay, etc., [it’s] increasingly deregulated even as more burden gets shifted to the driver in the truck and the idea of monitoring. And so essentially sort of redistributing the imposition of where the regulation happens. […]
And just in the last week or so a second episode, "What Truckers Already Know About the Future of Electronic Worker Surveillance", with a researcher that just published a book on the industry:
* https://omny.fm/shows/odd-lots/what-truckers-already-know-ab...
* https://podcasts.apple.com/us/podcast/what-truckers-already-...
> Long-haul truckers are the backbone of the American economy, transporting goods under grueling conditions and immense economic pressure. Truckers have long valued the day-to-day independence of their work, sharing a strong occupational identity rooted in a tradition of autonomy. Yet these workers increasingly find themselves under many watchful eyes. Data Driven examines how digital surveillance is upending life and work on the open road, and raises crucial questions about the role of data collection in broader systems of social control.
* https://press.princeton.edu/books/hardcover/9780691175300/da...
I have sympathy for anyone who spends all day locked in a cab with a robot that's always ready to punish for infractions. From the outside it appears that the robot cares about profit, not about safety.
It depends on factors like load weight, truck conditions, rail route, and a bunch more.