Self-Driving Trucks Will Create a Megaregion Economic Boom

Urban areas are richer and more productive. Research shows that doubling the population and increasing urban density boosts productivity by about 15%. Roman cities were connected by the distance someone could walk in one hour. Self-driving cars, trucks and buses will be able to eliminate traffic jams and then have steadily increasing safe operating speeds. The German Autobon has safe operation with people allowed to drive at any speed they feel is safe. This ends up grouping drivers into 80 mph, 100 mph and 120 mph groups. Self-driving vehicles and having some dedicated roads and highways would enable rapid deployment of fast inter-region logistics. This would be deployed far faster than the decades China has used for high-speed rail connections between and in large cities.

Self-driving trucks and cars with higher speed and no traffic jams will increase the population within one hour travel that can closely interact. This forecasted economic boom does not initially involve moving anyone from where they currently live or work. There could be later urban evolution where regions are reshaped to more effectively take advantage of transportation changes. The volume of business and speed of business will increase as the supply chain is sped up and regional interactions and transactions increase. All vision automation can be a few hundred dollars or less allowing for ground automation of Curiosity Rover sized vehicles or even smaller. I think ground delivery will have lower cost most of the time compared to drones. flying drones will have a niche for difficult terrain (many steps, hills, etc..).

The ways in which density is linked to productivity has been developed in a wide array of research projects. Six key impacts are discussed in this section:
* density allows a higher degree of specialization, increasing efficiency;
* reduced transport time and costs for products/goods/services from one stage to the next, or from producer to consumer, occurs in denser areas if the transport infrastructure is sufficient;
* increased density increases the prevalence of knowledge spillovers, increasing innovation
* density allows firms to have access to larger markets of suppliers (especially labor supply) and consumers, allowing competition to enhance the quality of inputs and outputs;
* efficiencies of scale are created in denser markets where suppliers are reaching more potential customers;
* reduced land take in denser areas allows more economic activity to take place on a fixed piece of land than less dense designs.

Dan Graham (2005b, 2006) examined the relationship between increased effective density (which takes into account time traveled between business units) and increased productivity across different industries. Graham found economy-wide urbanization elasticity (that is, the response of productivity to changes in density) is 0.125. This means that a 10% increase in effective density, holding all other factors constant, is associated with a 1.25% increase in productivity for firms in that area. Doubling the density of an area would result in a 12.5% increase in productivity. Others have noted a 15% increase in productivity from a doubling of a population within an area.

There are megaregions in the US, Europe, China and rest of the World. The regions tend to be 80 to 250 miles across.

There are about 60 million people in the North-Eastern USA (Boston, New York, DC regions). Instead of this being a loosely connected cluster of 4 metro-areas self-driving trucks could integrate the region. This would be a quadrupling of connected people and potentially a 25-30% boost in per capita productivity.

The top 29 global megaregions generate $29 trillion in GDP. If connectedness was boosted with self-driving vehicles for a 25-30% boost in productivity for all megaregions this would be a global boost of $8-10 trillion. Boosting the entire global economy ($130 trillion) would add $30-40 trillion. Some megaregions are relatively close, so there is the potential for larger population integrations. China has about 250 million people in the cities around Shanghai. This is a three doubling integration.

In the USA, the northeast megaregion is close to the Great Lakes region and the Toronto-Buffalo cluster. If all three could be more tightly integrated this would be boosting the productivity of 130 million people.

SOURCES – Visual Capitalist, Wikipedia
Written By Brian Wang,

56 thoughts on “Self-Driving Trucks Will Create a Megaregion Economic Boom”

  1. Interesting short story.

    The state supplied terrafoam housing and food would be an economically efficient solution, the coming homeless problem due to the masses no longer serving a useful function could be an expensive problem.

    The terrafoam housing system reminds me of the "Monolithic Dome" farms in Texas that serve as rentals units for unfortunates.

    NYC 2017, it cost on average $73,000 to provide emergency shelter to a family and $38,000 to provide emergency shelter to a single adult.

    It would cost ~$5k to house a family per year in a "Monolithic Dome".

  2. Except once on the minibus (for commuting), it can deliver you directly to your destination by making sure all passengers on board are headed to roughly the same area, not just the same direction or same destination. Even if you are last off, if there's just 4-6 on board, the drop-offs will only add ~5 minutes to your commute – more than made up for by eliminating heavy traffic.

  3. If combined with a market economy – i.e. where there's a UBI but people spend on whatever they want – it might work. You'd just be replacing human business planners with computer/AI business planners.

    Humans would still be in the loop, at least early on, to train the AIs on planning for new products, innovative supply chain or production or distribution methods, etc.

    Where it could go wrong would be governments assuming they could just take over all productive capital, rather than relying on humans to own and control companies that own the productive capital. Without profit incentives, the bureaucrats in charge of directing the AIs would likely tend to avoid risky innovation and favor cost/corner-cutting that makes them look good and has no visible short term problems but eventually leads to disasters. (E.g. putting off fixing/replacing basic infrastructure, paying only to patch it up.)

    OTOH, taxing an ultra-rich owner class to support a vast majority 'consumer class' with a UBI may create some bad incentives as well.

  4. Point to point deliveries. No waiting to compose the cars of a train. No changing modes of transport from rail to local trucks or vice versa.

    Rail is pretty good for hauling a lot of the same thing – coal, grain, oil, etc, because no one is waiting on a specific lump of coal or bushel of corn. And those things get used in bulk so you can run a rail line from bulk souces to bulk users.

  5. "No traffic jams" is yet to be proven or even attempted so far as I've heard?
    I.e. SD cars currently don't coordinate, so it's entirely possible that they'll jam up roads just as much as human driven cars.

    Though faster-than-human reaction times might tend to reduce shockwave effects, so maybe?

  6. I've always liked to live away from the big city but still within a mile's drive.
    There's clearly an economic advantage to people living within an hour.
    Is there an increased advantage to having people live within a half hour? 15 minutes?

    I work about 15 minutes away from where I work, but some of the people live 45 minutes or more away. The difference means I can go home for lunch but they can't. If there's an emergency at the house I can take care of it in a pinch.

    This may not add any measurable economic benefit but it sure adds a quality of life benefit. But maybe there's an associated economic benefit at scale that I'm not seeing?

  7. I've done both ways, true market trade offs all over. I found being able to read on the bus really changes a lot of time considerations. I think this idea does not force changes, may even make central living cheaper.

  8. I much prefer a city where things are close enough to walk or cycle, to one where everyone has to travel by car. ( My commute time to work varies between one and two minutes, depending on whether I get the green light. Return is rather longer – it's my uphill workout.)

  9. The more automation there is, the fewer paychecks there are to pay for services no matter how inexpensive they are. The more AI and automation the fewer options there are for labor, even high skilled. This is going to eventually lead to labor that a computer simply cannot do even with human shaped robots. We cannot all be prostitutes, and I never wanted to date older men even 35 years ago let alone now.
    In the past, the economy simply expanded and absorbed labor from more efficient industries into newer sectors. I am not sure how that is supposed to happen with robotics and AI. Social unrest could get ugly.

  10. An elevator becomes more efficient the more you pack into it as well. By all means lets continue the strategy of packing without limit always results in a more wonderful world. I think more and more people are coming to the realization that Bladerunner cities are not what they are looking for when an application of the population brakes is available to us. The world in general is backing off on increasing densification because we do not see it as improving our lives but making it increasingly stifling and stressful.

  11. You're not entitled to your own facts, but you're entitled to your own tastes, however self destructive they might be. Cities, objectively, do attract a lot of people.

    Mind, I think part of that is regulatory pressure, and were the regulations that incentivize centralization lifted, most people would probably prefer something more like a suburban lifestyle.

  12. What if dedicated roads were built for self driving trucks between cities?

    They would only need to be one lane in each direction – no passing or having to slow down for thr truck in front of you changing speed. All of the trucks would be set to the same speed limit of 50-60 mph. And trucks would run on them 24 hours per day.

    There wouldn't need to be any highway patrol or road signs etc. Just one lane in each direction with a few exits. The trucks could be spaced out far enough where if one has to slow down to exit, the trucks behind it could keep going 50-60 mph and not have to slow down.

    I think a lot of people in passenger cars would welcome not having to share the roads with trucks.

  13. This is a really dumb idea, but something tells me it might come true:

    If robot cars are limited to the speed limit and humans continue doing what they naturally do, which is drive over the speed limit, this will force governments to do one of two things, out of a sense of fairness (and intense lobbying by the car industry). Either they will have to ramp up enforcement of speed limits, which is not logistically or legally possible, or they'll allow robot cars to drive faster.

    At first, robot cars will be programmed to drive slightly over the speed limit, just like humans. But once we start down the line of thinking of "we can let the robots go faster because they're safer", then it's not much of a leap to allow them to drive even faster than an average human.

  14. But it's not an "everyone's right" situation. Especially after the events of 2020, it's perfectly obvious that cities are objectively bad places to be. Anyone who can't see that is just dumb.

    And this isn't just in modern times. At no point in recorded history was it ever safe to be in a city.

  15. A society where autonomous tractors grow the crops and autonomous trucks deliver every day goods to regions with hundreds of millions of people is a society where cyberwarfare becomes itself a weapon of mass destruction with capabilities higher than a nuclear strike.

  16. In theory, coordinated self-driving vehicles could do things like all the vehicles at a stop light accelerating at the same time, instead of each accelerating a human reaction time after the one in front visibly began moving. In theory, you could even skip the whole stoplight thing, and just have the vehicles continuously going through the intersection with each direction arranging to go between the gaps in the other direction.

    In theory.

    In practice, this almost completely eliminates any margin for dealing with mechanical failures, and renders the whole system incredibly vulnerable to hostile spoofing.

    Not to say you couldn't increase traffic density to some extent, the least competent drivers throw a lot of friction into the system, and some of the causes of traffic jams such as self perpetuating slowdowns lingering long after accidents, could be eliminated by tuning some response parameters.

  17. Based on USPS data, Florida population is already above the 2025 projection. I work in CRE here in FL. The big story in this state is “Orlampa” which is basically Orlando and Tampa growing toward each other into one city. I see permits and project proposals as part of my job. The amount of manufacturing, assembly, service, and distribution proposed for this area is staggering.

  18. Why are self-driving trucks better than rail corridors? Rail has been around for a long time, it's a mature technology and all the issues are known and ironed out. Self-driving trucks are as yet unrealized, and may have a host of problems associated with them.

  19. While I tend to share your opinion, I have to recognize that my distaste for cities isn't widely shared.

  20. "…Under the above set of assumptions (current values, future growth, and the reduction in value from the new tax), a decade from now each of the 250 million adults in America would get about $13,500 every year…"

    They're always the possibility of unintended negative incentives promoting undesirable actions by the "uncivilized" to increase their personal slice of the pie.

  21. Careful with the word "density" here. Fast trucks give the same city/size benefit with *less* actual density, that is, spread over a larger area, but same "effective density" as travel time decreases. Maintaining/increasing the actual density over a larger area now connected by transpo is even better!

  22. As others have pointed out here, densely packed automated vehicles, especially trucks, will only get you so far until there are diminishing returns, especially in pre-pandemic clogged cites like NYC (the pandemic and WFH has done more to reduce traffic than automated driving will ever do). And what about Last Foot Delivery? We are a very long way from robots smart and agile enough to unload a truck with packages due all over a neighborhood, and deliver them to a front desk with a doorman, or to individual homeowners. Buildings themselves would ideally have to be redesigned to allow conveyor belts to internal package rooms…where packages would still have to be scanned and sorted for resident pickup in large apartment buildings, or individual homes equipped with externally accessible lockboxes (third party package theft is a huge and growing problem). Trucks that drive themselves are the least of the problem in today's ecommerce world. It doesn't pay to employ building staff to unload packages delivered at odd, unpredictable times of the day, either. Human drivers will always have a job until that problem is solved.

  23. yea, I don't really get the part about fewer traffic jams. especially if those trucks are still sharing the road with human-driven-cars. is he talking about building entirely separate highway systems for the self-driving ones so they can go at 120mph? or banning all human-driven cars so the self-driving ones can speed up? Either of those would seem to be many many decades away, if ever.

    The only thing I can see reducing traffic is maybe a few less people will own their own cars since the self-driving "taxi's" will proliferate. they will proliferate because people with self-driving ones will rent them out so the "robots" can work as taxi's when their owners aren't using them, recouping some of the cost. Assuming riders don't trash the upholstery too badly. And there are enough robot taxi's to deal with the rush-hour "surge" where everyone is trying to use them at once.

    In the near term, I can envision robot-driven systems actually being even less efficient, so long as it is a mixed system with human drivers, since the robots will be stuck driving the human speed limit due to safey/liability reasons. Some skilled humans might be able to driven 10-15mph over the speed limit when traffic allows and getting places more efficiently, but robots will be prevented from using similar judgement, and be stuck at the programmed 60mph cap.

  24. Even further off, self-driving buses would never leave early. The drivers who leave early were obstructed at birth ("Get out of my way!"), likely, and the ones who are always late were induced at birth ("Wait for me!"), likely, unless of course there is traffic. Mix them, and there is really no bus schedule. Ridership would increase if this were fixed.

  25. It is often claimed that full self driving vehicles would coordinate among themselves (or be coordinated by a central system), and this would greatly increase the efficiency with which the roadways were utilized, thus eliminating or reducing traffic jams. I don't know whether there has ever been any analytical or empirical proof of that claim. I seem to recall some computer simulations of traffic that included such coordination of the actions of all the vehicles, and it did show that far more traffic could be handled before jams began to occur. But I don't know whether the same thing would occur in the real world.

  26. As I read the post, I was asking myself whether the claim that all centrally-planned economies fail because the central planners cannot possibly have or comprehend all the information needed to make the right decisions still applies when the central planners are computers running artificial intelligences that pull in just about any bit of data they want.

    I have a feeling that the first few attempts at implementing such central planning would fail, just as prior attempts have failed, because not all the information needed was available or was not taken into account. I don't have a feeling one way or the other whether allowing the AI to improve itself would eventually lead to central planning that works.

    And even if the AI can improve itself into something that works, would such a system be one that the humans would welcome, or would it be intolerable to most humans?

  27. I'm thrown by the assumption that driverless trucks equals no traffic jams. This seems an incredible reach.

  28. If you explode VMTs, even with perfect operation, road capacity is going to be a challenge. We will likely need a smarter uber pool like service with minibuses, particularly for these long-haul trips within a megaregion. It might be a bit like train stations where a personal vehicle brings you to a mobility hub where it matches you with other people going in the same direction to board a minibus that can speed off to another hub (or major destination like airport, stadium, etc.). Cities are going to choke on tobotaxis if they are all single occupant. We'll need road pricing to manage the traffic at peak times.

  29. Well yes, from the Manna story, that is one fork in the UBI future (managing an idiocracy phase while undermining reproductive rates to effectively exterminate the poor through the passage of time rather than just flat out executing them, to make ourselves feel better about forced depopulation), the other fork being hippie utopianism under an AI police panopticon. Though that ends up being more about being able to see the bars of your cage or not…

  30. There may be interesting pressures with full self driving that will alter the mix of vehicles in the logistics chain. With FSD, there may be more pressure for a larger number of single cargo container carrying sled vehicles as a logistics swarm doing more point-to-point work to more local distribution points, rather than traditional large 18 wheeler semis and larger megawarehouses (especially if container loading/unloading becomes more automated, moving from palleted goods to random box jumbles). With the potential for reduced overall length, we may see more trailer doubles/triples well, if not close range adhoc cooperative platooning.

    Either way, automated pallet unloading bots for local businesses become key, if one isn't carried by the truck, plus parcel carrying walkers or cargo drones deploying into a neighborhood from a local delivery vehicle (if not direct human pickup from that vehicle).

    But while there may be big economic growth from reduced logisitics costs and faster delivery, the economic costs from a huge number of unemployed drivers/delivery people is something that also needs to enter that growth calculus.

  31. What's "tasty" about cities? I'm not exaggerating when I say this: I have never felt comfortable in any city. Being in a city gives me anxiety. The bigger the city or the deeper into it I am, the more anxiety I feel. And it's not introversion, I'm actually pretty extroverted. It's more that cities are ugly, full of creepy people, homeless people, and a lack of any real social connection. Oh, and they're always one bad event from a riot breaking out.

  32. One of his suggestions

    "We could do something called the American Equity Fund. The American Equity Fund would be capitalized by taxing companies above a certain valuation 2.5% of their market value each year, payable in shares transferred to the fund, and by taxing 2.5% of the value of all privately-held land, payable in dollars.

    All citizens over 18 would get an annual distribution, in dollars and company shares, into their accounts. People would be entrusted to use the money however they needed or wanted—for better education, healthcare, housing, starting a company, whatever."

    I expect it will be spent mostly on beer and drugs, good solution for those that no longer serve a useful function.


  33. From Sam Altman's(OpenAI) blog :
    "…In the next five years, computer programs that can think will read legal documents and give medical advice. In the next decade, they will do assembly-line work and maybe even become companions. And in the decades after that, they will do almost everything, including making new scientific discoveries that will expand our concept of “everything.”This technological revolution is unstoppable. And a recursive loop of innovation, as these smart machines themselves help us make smarter machines, will accelerate the revolution’s pace. Three crucial consequences follow:"

    The consequence to the progress he sees in the next 5-25years is essentially socialism, it's hard to put some lipstick on that pig and call it something else.
    People that fetishize ignorance and stupidity aren't usually known for planning ahead, localized economic collapse is also a real possibility if the speed of these advances increase. We're at the beginning of the age where the labors of the poorly educated masses are no longer needed. Society will have to find a way to reduce their numbers or they will become a bigger festering problem than they are already.

  34. Going further down that road, the digital nomad life is the best cure for current obsessions by mostly insular people.
    The world is a big interesting place and is the antidote to all the delusional thinking that accumulates out in the sparse hinterlands.

  35. Somewhat off topic: I'm hoping that self-driving public transit will help with the housing crisis in California, by letting the people who have to be physically present live a bit further afield but still have shorter and less stressful commute. It might also provide some mitigation to the way that gentrification is breaking up communities. It's tough to choose between a) gaining a financial windfall by selling your house in south west Berkeley, but having to move to Tracey or Fairfield, and b) staying in the community where all your friends and relatives are. Easier, more frequent and less stressful transit between exurbs and the urban cores would ease some of that.

  36. urbanization, education, higher productivity, lower number of offsprings, etc… are all part of the same phenomenon.
    Btw, since you hinted at biology, when a species goes up in the food chain, it decreases the number of offsprings as a way to self-regulate its burden on the ecosystem. Critters at the bottom of the pyramid make thousands of offsprings, of which few survive; apex predators, who usually have a long infancy in which they learn to deal with a complex society, have very few offsprings. It's metabolically very costly to raise one orca cub compared to millions of krills. Now, it may sound horrible, but the rural proletariat surely takes less time to train (start working at 18 at latest) and settle (earlier marriage or first child) and have a higher number of offspring, on whom they invest relatively less. Upper class folks seem to be obsessed with education, those kids with 2 PhDs start working at 30 and aren't too much into changing diapers for a good decade after that; they consume a lot compared to proletarians, who typically work for them. How's that for a biology comparison..? This planet can handle only so many orcas. Now 5bn more people want cars, meat and A/C, can't nuke'em to stone age to keep our lifestyle, we got to adapt to survive. As the level of individual consumption rises it is inevitable that the number of offsprings decreases, until we get to unlock the resources in this solar system or at least learn to better manage what we have.

  37. Is economic output the only way to measure wealth? I grew up in a rural community out on the High Plains. We had abundant fresh food, great healthcare, great schools, community pride and involvement, low crime, inexpensive housing with all the modern amenities, strong personal relationships, no traffic congestion or long commutes, low stress levels, and plenty of outdoor recreation. And now, with the addition of high speed internet, great entertainment options accompanied by remote working…

    By the way, the education was great – I was recruited by and graduated from an Ivy League School (Brown University) – although neither parent graduated high school.

  38. We could house small families in the 24 hr rigs and they could be there for emergency driving etc. Joking. Or, am I?

  39. When given the perfect resources *other than* space, even rats know when there are too many rats. Does this keep uselessly large numbers down to starve/limit parasites and predators? Humans are in worse shape, as we are addicted, as you exactly correctly say, to the stuff that crowds us together, money and power, etc., found in cities. Something is amiss if healthy children is not the goal. How many? Pretty sure matriarchal clans are the answer.

  40. The main shipments are with 18W rigs then there is last mile and last foot delivery from local warehouses. Self driving works with smaller trucks, cars and 2-3-4 wheel automated delivery within a few blocks. All vision automation can be a few hundred dollars or less allowing for ground automation of Curiosity Rover sized vehicles or even smaller. I think ground delivery will have lower cost most of the time compared to drones. flying drones will have a niche for difficult terrain (many steps) Self-driving trucks and cars with higher speed and no traffic jams will increase the population within one hour travel that can closely interact. This forecasted economic boom does not initially involve moving anyone from where they currently live or work. There could be later urban evolution where regions are reshaped to more effectively take advantage of transportation changes. The volume of business and speed of business will increase as the supply chain is sped up and regional interactions and transactions increase.

  41. Self-driving trucks and cars with higher speed and no traffic jams will increase the population within one hour travel that can closely interact. This forecasted economic boom does not initially involve moving anyone from where they currently live or work. There could be later urban evolution where regions are reshaped to more effectively take advantage of transportation changes. The volume of business and speed of business will increase as the supply chain is sped up and regional interactions and transactions increase.

  42. Get Elon to put grids of tunnels under the cities for self-driving electric trucks. Then get all the trucks off the streets permanently. As soon as possible, do the same thing with cars (but different tunnels).

    Would make city life sooooo much less objectionable. Never mind the noise, and pollution, and never-ending crosswalks, a bit at a time we got used to the idea it was okay for tons of metal to go slinging around between all us soft squishies at relatively high speeds. It really shouldn't be.

  43. I have been working a high value job that is located in a city center but have not set foot there in a year. My daughter has been teaching High School in North Carolina from Mexico. Things are changing.

  44. Are we sure that self-driving 18W rigs are the way to go? I don't think you just take the existing tech, infrastructure, business model, and slap an autonomous system on it and say: "economic boom". What are the big costs – drivers, warehousing, parking, fuel, human controllers/ loaders… etc. It seems to me that smaller trucks, constantly driving and dropping off/ picking up at smaller venues and tight residential streets, all-day, all-night would be way more effective. A central AI distribution controller. Curbside dropboxes that can be loaded up autonomously off the truck and be accessed by either last-mile staff or consumers. If we are automating, we automate the whole vertical stack of manufacture-package-ship/out-delivery.

  45. They may have economies, but Liber is right: Throughout all of recorded history, cities have been population sinks, where people went to accomplish stuff, sure, but then die without having enough children to replace themselves. 

    That hasn't changed in the modern era. In fact, urbanization is probably a large part of why developed countries are suffering from a birth dearth. 

    You can't call an environment in which a species doesn't reproduce a good environment for that species: Cities are a deadly trap for humans. That the bait is tasty doesn't change that. If we're to survive we have to break our addiction to them.

  46. Densely populated areas have economies with high added-value industries, whose workers are typically better educated than low added-value rural workers. Education is the main cause in a change of behavior. In large datasets, the higher the education (often wrongly measured with the formal years of schooling… there are better proxies for functional education) of an individual, the less offrspings they'd have – and have them later in life too. Same for pretty much all the other behavior metrics (income, life expectancy, chance of violent accidents, chance of imprisonment, substance addiction, unwanted pregnancies, etc…), they all depend mostly on education.

  47. I don't think it is only density providing increased economic activity. Cities draw more extroverts, fast talkers and less homebodies. There is a lot of self-selection as to where people will chose to live. 
    If you cram a lot of people into a city who hate cities, it will get ugly.
    Large cities also attract financial sector companies, because they can afford to pay the rent. And because, the social amenities like symphonies, opera houses, theaters, fine art galleries, museums, zoos, marinas and such attract these high-end workers or their wives. It also helps to have other religious centers besides just Protestant and Catholic. Having a Synagogue, a Mosque, a Buddhist Temple, and a Hindu Temple go a log way to attracting highly skilled people from arround the World.

  48. Density also lowers birth rates. Due to the pandemic I think large cities in the US may get physically larger not denser. The amount of open land in the US is mind boggling. The US can go out and not up.

  49. In Space, you have rotating habs that are in 3D separated by 0 g open pathways, all designed rather than adapted to 2D planet surface. Now, there we have efficiency!

Comments are closed.