Tesla’s “Semi” Truck: Perspectives from the Cab and the Truck Stop and the Dock

By Lambert Strether of .

In this simple post I’m going to consider the Tesla Semi truck — which Tesla CEO Elon Musk announced November 16 to , along with statements that — as a vehicle, and not a financial vehicle[1]. (I also won’t address the role for which the Semi may be most appropriate: Line haul, short haul, port duty, etc., though I may address these supply chain issues at a later date.) First, I’ll look at the Semi’s cab and wheels; then, I’ll look at the Semi at the truck stop and on the dock. In each case I’ll rely heavily on the informed commentary of the NC commentariat, as well as the wisdom of Hurbert Horan. First, though, here’s a look at the thing:

OK, OK, it’s a little dystopian, but since Tesla released it, who am I to argue?

The Cab and the Wheels

Starting with the cab, I’ll requote this, from :

[W]e’ll start with the central seating position.

A second reason I picked the image above: It shows that central seating position clearly.

I don’t see how that helps a trucker. I already get “a commanding view of the road” in a traditional truck because I sit six feet above traffic. What I need is a commanding view of my own truck, which the central seating position compromises. The worst [1]blind spot in a tractor is next to the doors; in the Tesla Semi, I can’t lean over to see if there’s a Toyota Corolla camped out beside me.

I just hope I’m not in that Toyota Corolla, even if it is a robot car!

The central seating position hampers my commanding view when I need that view most: when I [2]back up. For any backing maneuver, I watch both sides of the trailer in my mirrors to make sure I don’t clobber anything, or I lean out of the truck to watch the trailer as I back. Being able to physically watch the trailer – not camera images on screens – can be the difference between making a clean back-up or making an insurance claim.

Furthermore, I can’t see around [3]trucks in front of me without pulling halfway into another lane. When I need to [4]exchange paperwork with the guard at a terminal, or the police, I can’t lean out the window to do so. Speaking of which, I have to believe one of the windows on the Tesla Semi [5]rolls down, but I can’t figure out which one. If, as it appears from the renderings, the windows only vent, well… that’s unacceptable.

I’ve helpfully numbered the driver use cases that seem not to have been considered in the design of the Tesla Semi’s cab; five ([5]) seems like rather a lot, particularly since I get the sense that the writer was just getting started.

NC commenter altandmain raises a second issue, again, visible from the Semi image: Clearance (), also visible in the photograph above:

Look at the front and rear axles. It was no doubt for the sake of aerodynamics, but a production version would have to have it raised or else the semi simply cannot operate as is. Particularly in [1]hilly terrain, even trucks today can get stuck on the crest of a very steep hill (it does happen in the industry). Very simply put, the production version cannot be as aerodynamic.

[2]Snow will also be an issue with low clearance. This might be a big problem in Canada, the northern US, Alaska, and northern Europe. Russia too, if Tesla ever intends to sell to the Russians. Also, notice that the rear wheels are concealed. This could be an issue [3]if the vehicle is stuck in the snow (ever see a car get stuck in the snow)? Trucks have more ground pressure and as a result, are at a larger risk for “getting stuck”….

There must also be the ability to attach [4]moose bumpers. That is necessary in Canada and northern US. I did not see anything on the images that would imply that a bumper provision was available.

Here, too, I have helpfully numbered the use cases that the Tesla Semi’s designers ignored; four (4) also seems like rather a lot. No snow in Fremont, California, eh? Or moose, for that matter.

Now, the suggestion has been made that getting us lost in the weeds like I just did is a clever scheme by Tesla. Hubert writes:

Objective is to get people discussing minor design/implementation details in order to convince people that these vehicles are inevitable and only a few years away from mass production and widespread acceptance, and to distract from the fact that the underlying technology doesn’t exist, and even if it did the underlying economics are abysmal.

Plausible, but to me, this is a variation of The Eternal Question in DC: “Stupid, or evil?”[2]. And frankly, I’d consider going with stupid. Tesla is a creature of Silicon Valley, and there are plenty of software types who’d try — and even succeed! I’ve watched it happen! — to take a wireframe that “looked cool” all the way through to production while ignoring any and all User Interface/User Experiece inputs, out of sheer arrogance and indifference to the needs of others. And when I look at all the obvious use cases Tesla missed — nine really obvious ones, after a pretty shallow examination — it looks like one of those software types was in charge of this project.[3] (Lots of these types work on the Internet of Things.)

The Truck Stop and the Dock

One thing I’ve noticed about trucks is that they do a lot of parking and backing up and maneuvering around obstacles. That makes mirrors very important. The Semi doesn’t have mirrors. It has cameras that display on screens. From fan site :

Because the Tesla Semi doesn’t have any side mirrors, video from these cameras likely provide the driver with a birds eye view of both sides of the truck as seen from the interior touchscreen displays. Kman also discovered a row of cameras mounted below and above the massive windshield, and even more cameras discreetly mounted within Tesla Semi’s headlight assembly.

NC commenter Edward E writes on the limitations of cameras;

Some docks are easy to back into and the cameras would be fine for those. But a large number of docks are extremely challenging to back into. Lighting at night causes glare and cameras would have trouble with that, even with mirrors you occasionally have trouble. Really difficult docks are the covered ones where you must back into a dark building from the sunny outdoors and hit a dock while narrowly missing poles, walls and doors. The only way to see anything is to get out and look and then stick your head out the window and look straight back while backing in. The Tesla appears useless for that. Warehouses in underground caves are especially difficult in this regard and you may have to back up quite a distance around obstacles like rock pillars. I could tell you some stories about long distance backing a semi, sometimes I do it from the highway up a half mile of twisting County road to my driveway.

Altandmain makes an additional point. No backups:

one issue I see is that there are no manual backups. A camera can be covered by mud. If you own a car and you have a backup camera on your car, you have some idea of the limitations. There should be mirrors. Otherwise, the cameras don’t always work in all conditions. Plus the camera can die or face other problems.

No backups? For a computer system? Kidding, right?

And then, besides the cameras, there are those screens (Indignant programmer: “But I work with screens all the time!”) :

Another reason to have physical mirrors: so I can turn off, or turn down, the two giant screens in the cab (screens which, by the way, hinder my view of the corners of my truck). The light required to provide a useful camera image at night would kill my eyes during a full drive shift. Doing an 11-hour stint in a dark cockpit in the glow of large digital screens only works in anime and “Battlestar Galactica.” I had one computer in both trucks I drove, and unless I was using it, I turned the screen off.

“Anime and ‘Battlestar Galactica.'” Ouch!

Conclusion

In concluding, a word from NC commenter XXYY:

I gained a tremendous respect for people who drive trucks for a living. It’s difficult and demanding work, and there are great dangers in driving an 80,000 lb vehicle in heavy traffic, rain, snow, and city streets. The pay is low, the stress high, and long-haul drivers are frequently away from home for weeks at a time. Despite all this, the truckers I met had tremendous professionalism and pride in their work, and saw themselves as part of a collective effort to get everyone (not just their own load) where they are going quickly and safely.

Those are Musk’s users, and if the Tesla is to be a real product — instead of a bright shiny object[4] dangled before the Wall Street touts analysts to detract from , not that I’m cynical — he could start by listening to them. If you hose away all the technobabblish froth, and consider the truck as a truck, it becomes readily apparent he hasn’t. I’m sure I’ll have more to say on this topic in the future, especially on larger issues like batteries, charging, and the supply chain generally, but I thought it was important to kick the Semi’s tires first. As usual, comments from drivers and allied professions most welcome!

NOTES

[1] Though ; Musk and company could well have made that there first concern.

[2] Yes, I know it’s both/and, not either/or.

[3] That, or Musk thinks he’s Steve Jobs, except as “thin” is to Jobs, “aerodynamic” is to Musk.

[4] :

It’s probably best to view Thursday night’s truck-and-pony show Hawthorne as a bit of savvy marketing move, a down payment on the future of transport. “Given the production issues around Tesla and its significant cash burn, it might be easy to discount this announcement as more marketing than substance, but I expect Tesla will build this truck eventually,” said Michael Ramsey, an analyst at Gartner. “If the future is electric — and it likely will be at some point — then laying the groundwork now is not a waste of time.”

I’d expect more puffery from Gartner. Or perhaps that is the puffery?

Print Friendly, PDF & Email
This entry was posted in Environment, Guest Post, Infrastructure, Ridiculously obvious scams on by .

About Lambert Strether

Readers, I have had a correspondent characterize my views as realistic cynical. Let me briefly explain them. I believe in universal programs that provide concrete material benefits, especially to the working class. Medicare for All is the prime example, but tuition-free college and a Post Office Bank also fall under this heading. So do a Jobs Guarantee and a Debt Jubilee. Clearly, neither liberal Democrats nor conservative Republicans can deliver on such programs, because the two are different flavors of neoliberalism (“Because markets”). I don’t much care about the “ism” that delivers the benefits, although whichever one does have to put common humanity first, as opposed to markets. Could be a second FDR saving capitalism, democratic socialism leashing and collaring it, or communism razing it. I don’t much care, as long as the benefits are delivered. To me, the key issue — and this is why Medicare for All is always first with me — is the tens of thousands of excess “deaths from despair,” as described by the Case-Deaton study, and other recent studies. That enormous body count makes Medicare for All, at the very least, a moral and strategic imperative. And that level of suffering and organic damage makes the concerns of identity politics — even the worthy fight to help the refugees Bush, Obama, and Clinton’s wars created — bright shiny objects by comparison. Hence my frustration with the news flow — currently in my view the swirling intersection of two, separate Shock Doctrine campaigns, one by the Administration, and the other by out-of-power liberals and their allies in the State and in the press — a news flow that constantly forces me to focus on matters that I regard as of secondary importance to the excess deaths. What kind of political economy is it that halts or even reverses the increases in life expectancy that civilized societies have achieved? I am also very hopeful that the continuing destruction of both party establishments will open the space for voices supporting programs similar to those I have listed; let’s call such voices “the left.” Volatility creates opportunity, especially if the Democrat establishment, which puts markets first and opposes all such programs, isn’t allowed to get back into the saddle. Eyes on the prize! I love the tactical level, and secretly love even the horse race, since I’ve been blogging about it daily for fourteen years, but everything I write has this perspective at the back of it.

87 comments

  1. AdamCoppola

    Great summary. Ive seen a lot of articles telling me I should be afraid of autonomous vehicles. I also s*** myself when I saw the cgi video of this bully steaming past a sedan on a thin two lane road.
    Maybe ill spend less leisure time on the road, leave it for work.
    I mean to pose neurotic question: could they be scaring me on purpose?

  2. MW

    While it may not be perfect, I applaud Tesla for trying to push boundaries and move us away from fossil fuels. Even if Tesla fails as a company, they are pushing the traditional car companies to offer electric cars at a much faster rate than anyone expected. I have a sister in law in Texas whose family made their fortune in the oil business. She is considering buying a Model X because it is the new status symbol. Well done Tesla.

    1. Lambert Strether Post author

      like . For some, that’s a problem. I don’t applaud “pushing” that sort of “boundary” (although its common with Silicon Valley firms; see Amazon).

      I haven’t researched how electric vehicles net out environmentally.

      1. Buzz

        Dear Lambert, if you have “proof” that Tesla workers are being treated like garbage, you need to show us. It is NOT “common” in Silicon Valley to treat people like garbage. My wife worked at Apple. She commuted by train to the valley. She was picked up at the train by Apple, free of charge. They provide free child daycare. They have an extensive cafeteria where employees eat very well at low rates. Their wages are above average and “flex hours” allow employees to arrive any time in a wide window. Maybe you were referring to another Silicone Valley since Amazon’s headquarters is in Washington State. Please try to do more research before you rush to publish.It makes you appear to be sort of “pushing something”. P.S. Tesla is not in Silicon Valley. It’s in the east bay, Fremont to be exact.

        1. Lambert Strether Post author

          Do consider reading the two links I supplied; feel free to take issue with them. After that, consider getting your knee seen to.

          * * *

          On “Silicon Valley,” let’s not be tediously literal-minded. :

          “Silicon Valley” has come to have two definitions: a geographic one, referring to Santa Clara County, and a metonymical one, referring to all high-tech businesses in the Bay Area or even in the United States.

          Here’s the definition of . I hope this helps.

  3. voteforno6

    I mentioned this vehicle with a relative, who has some experience with trucking (as do others, actually). I brought up the central seating position, and he had the exact same reaction. It’s almost like Tesla is not concerned with testing this out with people who have practical experience with trucking.

    1. cocomaan

      I am not kidding, or proud, when I say that I had the idea for center seating in a car when I was five years old. That doesn’t say much about this truck.

      The only cars that I can think of with center seating are racing vehicles like the Mclaren F1.

      1. rusti

        Volvo Bus has the driver in the center position for the (but not the hybrids). But there are no backing maneuvers in traffic for those buses.

    2. Lambert Strether Post author

      I’ve been thinking about this, and if I invert the bugs into features, it means the “Semi” is designed for (future) automation, and the driver is sitting in the center to handle any emergencies, so ultimately we’re looking at a robot truck. (Which is still dumb, because that Toyota is an emergency.)

      But it’s really not clear to me that robot trucks are a thing, any more than robot cars, if we look at the use cases readers keep supplying. Of course, the object could be that every business not accessible by robot truck should wither and die….

      1. rd

        Everybody is focused on visual mirrors and cameras. However, if this is to be semi-autonomous or autonomous, then other sensors like radar will be necessary and may be the justification for the center seating. The radar should provide continuous coverage around the truck in all but the worst weather conditions. So the camera images would provide a visual depiction of the object, but the radar would tell the driver where it is and should eliminate blind spots. If needed, they could even do IR sensors or night vision cameras as well.

        1. redleg

          If any vehicle is meant to be autonomous, a (mostly) closed system like rail should be first in line for it. Robot vehicles on a road network (open system) will kill a lot of people, directly and indirectly, through acts of commission and omission.

  4. Tyronius

    The only way a central driving position makes sense is if Tesla plans to make the vehicle autonomous from the very beginning. That way the ‘driver’ is just another passenger, conveniently redundant in the glorious digital future! All the issues mentioned about cameras would apply to self driving as well. Why do I suddenly feel even less comfortable with the prospect of sharing the road or loading dock with self driving trucks in bad weather?

    1. Summer

      Looking at the issues pointed out, it’s not just the truck driver expected to be eliminated from the road.

      “Why do I suddenly feel even less comfortable with the prospect of sharing the road or loading dock with self driving trucks in bad weather?”
      These appear to be designs that are expected to work with no human drivers on the road, with ideal conditions being other autonomous vehicles on the roads (with infrastructure changed to adapt to them).

      They’ll weed out human drivers through insurance costs…

      1. Lambert Strether Post author

        > with infrastructure changed to adapt to them)

        So, when it’s time, that’s where all the infrastructure money will go! (as opposed to, say, schools, or public wifi, or…)

        > They’ll weed out human drivers through insurance costs

        Of course they will (though I don’t know how you do actuarial calculations for robot AIs. Perhaps they’ll just jigger the numbers for the desired outcome…)

      2. rd

        I have seen quite a few 18-wheelers buried deep in the woods along interstates in the northeast over the past year. I think the human drivers are working very long hours and literally falling asleep at the wheel. That is the only explanation for 18-wheelers that have gone 100 feet off the road along a flat, straight interstate and not stopped until only the rear door of the trailer was sticking out form the trees. In August, I had one four-hour drive where I saw three of these. It was pretty scary because there were a lot of trucks around me.

  5. Robert McGregor

    It’s one thing to rush to market a concept “with bugs.” It’s another to rush to market a concept that’s not even designed to work (center seating; no mirrors; low bottom-clearance) !

    1. Lambert Strether Post author

      > low bottom-clearance

      See, when the truck gets stuck in the snow, all the little robots will jump out and lift the truck up with their tiny little metal hands. Same with waving away the moose.

  6. visitor

    There are also fundamental technical issues with the announced truck: its entire power subsystem is simply unfeasible with current technology. The article gives an explanation.

  7. California Bob

    “No snow in Fremont, California, eh? Or moose, for that matter.”

    Nope, but we’re gonna have elk before too long:

  8. autoagri

    And when I look at all the obvious use cases Tesla missed — nine really obvious ones, after a pretty shallow examination — it looks like one of those software types was in charge of this project.

    Actually Jerome Guillen, formerly of Daimler Trucks North America, led development of the Tesla Semi. Previously he led development of DTNA’s popular Freightliner Cascadia line of trucks. Shallow examination indeed.

    I have no interest in Tesla’s financial success — arguably the opposite — but I have taken a test drive in a Tesla Semi, talked to some of its developers and prospective customers, experienced its acceleration and turns… and I consider it to be a commercially viable product, perhaps as soon as 2019. We’ll see.

    I’ve been a loyal reader of NC, including the comments, for much of the last 10 years. I greatly enjoyed the Hubert Horan series of articles on Uber’s economics; virtually of his points rang true with respect to my knowledge as a practitioner in the ground vehicle technology sector. But NC’s recent expedition into technology-based criticism of Tesla and vehicle automation strikes me as overreach. Strong ethical and economic criticism of these subjects is ripe for the making — and would be consistent with my reading of NC’s general editorial perspective and leadership — and yet the temptation to deem oneself more knowing than scores of technical practitioners is apparently what’s driving the sector coverage most prominent of late. A few related points:

    – automated and electric vehicles are designed for specific use cases

    – the Tesla Semi’s center seating may indeed be gimmicky but still practical in many use cases; mirrors can certainly be added to expand the scope of viable use cases, with a small sacrifice in aerodynamics

    – to suggest that any vehicle as a whole is somehow not “real” because its prototype interior design seems gimmicky or otherwise flawed is to throw the baby out with the bathwater

    – to suggest that a Tesla vehicle specifically is not “real” because Elon Musk is an arrogant and/or desperate CEO is to ignore a history of overpromising but eventually delivering; even in bankruptcy, the vehicle designs will remain

    – to be reflexively against partially-automated electric vehicles powered by solar (see Tesla Semi charging station plans) and promulgated by Silicon Valley technologists is fine, and maybe wise if it coincides with opposition to resource-intensive human industry in general, but it would be more interesting at least to this reader to see that vision evaluated against the alternatives, and not just damned out of skepticism

    – finally, to cherrypick the members of the commentariat you say you rely on is off-putting and potentially a disservice

    1. Mike

      Well said. Although it is certainly possible that many of Musk’s ideas will not come to fruition in their current form it seems likely to me that he is moving the mark, as noted above, toward a good future. IMHO the carping about the inefficiencies experienced by innovators – many false starts and wrong turns – society never progresses without them.

      1. Lambert Strether Post author

        Musk is an extremely innovative marketer, that’s true. However, I don’t think that “innovation” (the Juicero) or “progress” (Fukushima) are concepts that should be accepted unexamined.

    2. JCC

      I think you make some good points.

      One of the things that I’ve been reading quite a bit lately, beside your observations, is that these types of commercial vehicles are not intended to be used for “last mile” deliveries. The odds are very strong that there would be outskirt warehousing built that would take into consideration the negatives like loading dock access, etc.

      Warehouses are relatively inexpensive to build, Amazon can have one up, running and fully functional within a year without batting an eyelash.

      I don’t find “autonomous” vehicles a good thing for a number of reasons (many of them covered here), but they will be everywhere soon enough at some level of autonomy, and just like infrastructure being built and/or modified for present day cars and trucks, future infrastructure, as it is built out, will take commercial autonomous vehicles and their requirements/limitations into consideration, particularly considering the strong drive in our times to reduce both human labor and its cost.

      It may not be Tesla, but BAE, Mercedes, Volvo,and others are already well on the road to autonomous, electric/hybrid commercial vehicles.

    3. voteforno6

      Just out of curiosity, in what circumstances is the center seating position practical? The column referenced in the above article had a very pointed critique of that feature, based on real-world experience.

      1. Lambert Strether Post author

        If the model is that the driver will just sit on the throne, waiting for the robots to tell him there’s an emergency, I think people doing human factors analysis for aircraft pilots might wish to have a word.

    4. bob

      ” I consider it to be a commercially viable product, perhaps as soon as 2019. We’ll see.”

      So, you have seen the price then? Or, is “commercially viable” now shorthand for “it moves”.

      1. Lambert Strether Post author

        > a commercially viable product

        Companies that don’t have to make a profit have a different idea of what’s “commercially viable” than the rest of us. Trucking companies have to make a profit, and in a low-margin business.

    5. The Rev Kev

      I would hardly call “NC’s recent expedition into technology-based criticism of Tesla and vehicle automation” being a case of overreach. Remember, one day we may be seeing these things come barrelling up behind us in our rear-view mirrors like something out of “Duel” so if people take a closer look at Ford or Toyota, then why should Tesla get a free pass? Tesla cops a lot of flak for its awful treatment of the people that work for it and not listening to them so if people, especially truckers, are looking a a concept design (probably with input from the marketing droids) and saying “not gunna work” as is, maybe it is time to listen to the people that will actually have to deal with these things. There are other problems external to this design as well.
      Not long ago there was a story of some hapless city in the US where one of those autonomous cars was being tested out. It wasn’t working out so well for the car as the roads had been run down and lines were not being repainted so the car was having difficulty recognizing where it was supposed to go. The car exec was complaining bitterly to the city officials how they should be repairing & painting more. Now I am not going to go into the aspect of how corporations have rigged things where they are putting in little or no money into taxes leading to budget shortfalls across the nation – and thus cutbacks into such things as road maintenance (that is another rant for another time) but how are these you-beaut trucks supposed to navigate on roads that are being increasingly being run down? California Bob brought up the valid point of crossing wildlife as well.
      Also, here in Australia a coupla years ago the trucking industry was busted for forcing their drivers to exceed the speed limits to get cargoes delivered. The trucking companies were regarding the fines as a cost-of-doing-business but the drivers were dosing themselves up to the gills on drugs to stay awake and alert. Would these autonomous trucks be programed to have trucks with special cargo (read Amazon) exceed speed limits to meet deadlines? It’s all fun and profit until the first autonomous truck wipes out a school bus. The only solution that I can see to help avoid collisions with these things would be to force all cars to have transponders like aircraft so that these trucks know who and what is around them. And even then with aircraft you still get midairs.

      1. redleg

        RevKev- I apologize in advance for this reply…

        Transponders… that can then be used for tracking a vehicle 1440 minutes a day. What could go wrong?
        Who would pay to equip transponders on existing vehicles? People with few resources drive the cheapest used cars because they have to. (And if you live anywhere outside of a metropolitan area, you have to have a car.)
        What about bikes- will they get transponders too? Joggers? Raccoons?

        Autonomous vehicles + big pharma + cameras! + transponders for everything starts to look a lot like THX1138.

    6. Lambert Strether Post author

      > temptation to deem oneself more knowing than scores of technical practitioners

      I’m uncertain how looking to actual truck drivers means that I “deem myself more knowing.” Unless you don’t regard truck drivers as “technical practitioners,” of course (but see the Conclusion) which would drive home my point on over-valuing software engineering rather neatly.

      > Acutally, Guillen, formerly of Daimler Trucks North America, led development of the Tesla Semi. Previously he led development of DTNA’s popular Freightliner Cascadia line of trucks. Shallow examination indeed.

      Actually, I wrote:

      And when I look at all the obvious use cases Tesla missed — nine really obvious ones, after a pretty shallow examination — it looks like one of those software types was in charge of this project.[3]

      You are familiar with the American idiom “looks like”? Surely — speaking of shallow — it’s possible that even the most gifted mechanical engineer could produce results that “look like” they were produced by an arrogant and ignorant software engineer? Much in the same way that a master chef could produce a dish that “looks like” it came out of a reheated plastic bag?

      * * *

      To your other points:

      > – automated and electric vehicles are designed for specific use cases

      It would be helpful if you would adduce the use cases for which the “Semi” is actually useful, as designed. I agree that you can simply turn bugs into requirements; no snow, no backing up, and so forth…

      > – the Tesla Semi’s center seating may indeed be gimmicky but still practical in many use cases; mirrors can certainly be added to expand the scope of viable use cases, with a small sacrifice in aerodynamics

      Yes, the product as released can be redesigned to eliminate all the sloppiness and stupidity. Is that your point? The real issue is why the design was released as it was. They’re actually trying to market the thing, as you know; it’s not a prototype or a mockup or an alpha version.

      > – to suggest that any vehicle as a whole is somehow not “real” because its prototype interior design seems gimmicky or otherwise flawed is to throw the baby out with the bathwater

      If you think adequate clearance for snow is a “gimmick,” I suggest you consult with actual drivers.

      > – to suggest that a Tesla vehicle specifically is not “real” because Elon Musk is an arrogant and/or desperate CEO is to ignore a history of overpromising but eventually delivering; even in bankruptcy, the vehicle designs will remain

      Er, no. “To suggest A is to ignore B” just is not so. The vehicle could well be (A) not “real” as serious and knowledgeable people in the field suggest and Tesla could (B) ultimately deliver something. As for eventually delivering, your claim is (for example) that because Lockheed delivered the P-38 (a prop-driven fighter), it also would have delivered the Lockheed L-2000 (a jet supersonic transport). I think it makes more sense to make concrete evaluations of products, rather than resort to vague hand-waving about company history. As for the designs, why assume they’re any good?

      > – to be reflexively against partially-automated electric vehicles powered by solar (see Tesla Semi charging station plans) and promulgated by Silicon Valley technologists is fine, and maybe wise if it coincides with opposition to resource-intensive human industry in general, but it would be more interesting at least to this reader to see that vision evaluated against the alternatives, and not just damned out of skepticism

      Ah, yes, the old “the author wrote a book about seagulls, but I’ve got to give it low marks because I would rather it were about penguins.” Which is all fine, and other posts will be forthcoming. (The question begging here is: “just damned out of skepticism,” and “reflexively against,” and the answer to the first is that I adduce data. The answer to the second is that it’s an obvious ad hominem.)

      1. autoagri

        Thanks for the extensive reply.

        I appreciate your willingness to accept the original critical comment.

        The real issue is why the design was released as it was.

        I’d start with fuel savings / range extension. As presented*, the Tesla Semi has a drag coefficient of 0.36, compared to a Cd of ~0.6-1.0 of typical tractor-trailers (e.g. Peterbilt’s aerodynamic 579 with a box trailer at the low end). This is an important design factor because about 2/3 of tractor-trailers’ fuel consumption at highway speeds goes to overcoming aerodynamic drag. (* As presented: including with just-above-grade skirting that, if it’s inflexible, would need to be removed to avoid damage at railroad crossings, etc.) I’d suppose — all I can do — that the greater curvature of the windshield and the inset from the wheelbase at the level where the driver sits, which contribute to drag reduction, both support moving the driver’s position towards the center of the cab. Once the designers elected to use video screens — not just as an alternative to mirrors, but also to cover blind spots that are problematic even with mirrors — then they may have figured that center positioning would create adequate and well-balanced space for large-screen displays on both sides of the driver. I say all this to answer your question. I also acknowledge the trade-offs of unfamiliarity and problem-causing for drivers, and uncertain (at least to us) performance in heavy weather. I think though that Tesla is probably aware of these tradeoffs yet is banking on (has reason to believe?) that they will be mitigated by driver training and counteracted by coverage of blind spots left uncovered by mirrors alone (notable that there is no mention in the AutoBlog piece that tractor mirrors leave blind spots)

        If you think adequate clearance for snow is a “gimmick,” I suggest you consult with actual drivers.

        I don’t think clearance is a gimmick; I was plainly (but unfairly?) characterizing your view of the interior design of the vehicle. Indeed I have spoken with many a driver about clearance with respect to trailer skirting (again, this is my business!). Ramps and railroad crossings pose problems even where snow does not. Thermoplastic skirting can be quite flexible and hold up to a lot of wear and tear. Flexibility though not be of any help when turning in the snow. I haven’t had the chance to look closely at Tesla’s materials, but agreed that they look precariously low to the ground. Some skirting is readily removable (5-15 minutes). Again I don’t know about Tesla’s.

        I think it makes more sense to make concrete evaluations of products, rather than resort to vague hand-waving about company history. As for the designs, why assume they’re any good?

        My point is that this evaluation does not seem well-hardened. As for company history, you recount only back to last month’s Model S news. Is that also hand-waving or something else? In any case, I agree that past results do not guarantee future outcomes.

        It would be helpful if you would adduce the use cases for which the “Semi” is actually useful.

        – Short- to regional-haul / out-and-back runs with ranges up to ~400 miles (no sleeper option yet presented); charging station locations will be a big factor in identifying suitable regional haul routes
        – Urban areas with emission-mitigation rules/goals
        – Confident, experienced drivers interested in learning / using new equipment (same approach used with adoption of forward collision avoidance systems); converse, don’t ask drivers who aren’t comfortable using screens for docking to do so. These parameters are a function of driver interest / acceptance and training, not absolute good or bad fits for the vehicle.
        – Since I don’t know how flexible the skirting is or whether it’s removable, I could hardly guess at whether grade variance or snow will be significant limitations
        – For peak vehicle efficiency, private / dedicated routes with limited or no trailer switching to hold together the tractor-trailer aero package
        – Unless and until drivers are comfortable operating the truck in heavy weather, urban driving, etc., don’t adopt for routes where these conditions are common

        I’m sure that doesn’t cover it all but it’s an honest effort. I don’t think the vehicle is perfect. It’s just better in significant ways, based on my own adducement of data, than any truck I’ve ever observed — in fact ridden in. And if I were a blind technology enthusiast, than it would be surpassingly strange for me to spend so much of my leisure time at this site. More data to adduce.

        In closing I’d like to add again that I appreciate your thoughtful response, as well as the perspectives of others who have addressed issues including labor practices and sharing the road with robo-trucks; all important, and not my focus in these comments about technology coverage.

        1. Robert McGregor

          > converse, don’t ask drivers who aren’t comfortable using screens for docking to do so

          Since you have some expertise in this field, can you some current examples in the transportation world where backing-up or any parking is done with ONLY screens? Some cars and trucks have backup screens, but this only as supplemental assist. I don’t know about the aviation world, but the airliners I have seen backed-up do so with the “human assist” of guys walking behind waving orange “stick lights!” Ocean liners have “navigation cameras,” but I don’t think captains have ever tried to dock their ocean liner from their windowless quarters with no other human assist. And by the way, Autoagri, have you ever driven a big truck? Unless you have access to some interesting prototypes, and are a “truck test driver, I don’t think you have ever backed up a big truck using just a screen.

          1. autoagri

            Great question. Here’s how I replied to an assertion in the original AutoBlog article excerpted in the 11/21 Water Cooler (one of a few comments that were not relied upon in this piece), which read, “Being able to physically watch the trailer – not camera images on screens – can be the difference between making a clean back-up or making an insurance claim”:

            This is sensible and true… as of now, when [] drivers generally have zero experience relying on screens to back up tractor-trailers. Doesn’t mean though that drivers won’t adapt to using the twin 15″ screens that flank the wheel in a Tesla Semi.

            Back-up cameras in passenger vehicles are an inadequate analog, because they’re only used to gauge 1 or sometimes 2 dimensions in a straight vehicle, whereas truck drivers are often dealing with 2 or even 3 in an articulated vehicle. But at the same time, unlike most passenger-car drivers, truck drivers are professionals, who are likely capable of learning to rely on large-screen video to back up slowly and safely. Finally, any worry at all along these lines is only valid until back-ups/docking are automated; that capability has been demonstrated in test environments but probably won’t be commercial for a couple-few years.

            So yes, the center seating would be — as Lambert notes — “[v]ery important” to a driver who today is plopped in the cab of a Tesla Semi and told to go about his or her business, but I believe it’s not an important safety issue [edit: it would be fair to add here, “in most on-highway tractor-trailer applications”] provided some training and/or further advancement in truck automation.

            Just as passenger-vehicle back-up cameras make for an imperfect analogy, I’d argue that so too do your examples of airplanes and ocean liners, which have different geometries and, in the case of the latter, are exposed to more dynamic environmental conditions (e.g. waves!).

            But at bottom I readily grant limited precedent; if there were better precedent than this wouldn’t be as hot a topic! Only recently have the price of off-the-shelf sensors and reliability of related systems engineering made the screen concept worth considering. It’s that factor combined with a belief in human learning capacity that informs my favorable opinion on the commercially viable aspect of Tesla’s design.

            As for your “by the way”, since you asked, I haven’t done more than roll around a well-buffered controlled environment in an auto-tranny tractor-trailer. My certified professional driving experience is limited to postal vehicles and 15-passenger vans. My articulated-vehicle back-up experience goes no further than with 3/4 ton pick-ups and a 20-foot trailer carrying heavy loads on- and off-road on a variety of grades with plenty of tight multi-point turnarounds In those circumstances I have appreciated mirrors and a driver’s seat position that enables me to put my head out the window. After all there were no alternatives, and even if there were any, like video screens, I would have preferred to go with what I know — not least because I haven’t trained do operate the vehicle any other way.

            So no, I don’t hold a CDL. In working with CDL-holding professional drivers, including in the development of training instructions for handling aerodynamic devices and using low-level automated driving systems, I have learned that, off the bat, many are receptive to new vehicle features and many are not. Forces in the industry being what they are, at least some of the latter reluctantly come around. Finally, I put great stock in their judgment — especially once they’ve tooled around a bit — about a feature’s limitations, i.e. when it seems helpful, when it feels unsafe, when it’s an annoyance.

            Much as my counterpoints to Lambert’s perspective might be construed in this venue as weighing lightly or even invalidating the voice of the driver, that’s bullsquat. I chimed in to add my own voice, which is pretty well spent on the topic of whether electric and automated vehicle technology assessment here is biased by concern, anger, skepticism, etc. regarding Silicon Valley PR practices, social effects of robo-trucks, and so on and so forth.

            Please take it or leave it. I’ll go back to being a lurker with high regard for NC’s overall body of work.

        2. redleg

          I would dare you to drive a truck using screens instead of mirrors while simultaneously looking out the windscreen but that would endanger to many people.
          Try taking a picture with your phone in bright sunshine. Then try that on your bicycle going down a hill.

        3. bob

          “I’d start with fuel savings / range extension. As presented*, the Tesla Semi has a drag coefficient of 0.36, compared to a Cd of ~0.6-1.0 of typical tractor-trailers (e.g. Peterbilt’s aerodynamic 579 with a box trailer at the low end).”

          Why would going “electric” have anything to do with the drag? In other words, why haven’t the current semi manufactures changed to the more aerodynamic shape?

          Tesla went and slimmed down the cab, and made the skirting lower. Those two things alone could result in the claimed reduction in drag.

          But, we can make many, many more sexy words than that. Let’s see how “sliming the cab” translates in sales speak

          ” I’d suppose — all I can do — that the greater curvature of the windshield and the inset from the wheelbase at the level where the driver sits, which contribute to drag reduction, both support moving the driver’s position towards the center of the cab.”

          And lowering the skirting-

          “I don’t think clearance is a gimmick; I was plainly (but unfairly?) characterizing your view of the interior design of the vehicle. Indeed I have spoken with many a driver about clearance with respect to trailer skirting (again, this is my business!). Ramps and railroad crossings pose problems even where snow does not. Thermoplastic skirting can be quite flexible and hold up to a lot of wear and tear. Flexibility though not be of any help when turning in the snow. I haven’t had the chance to look closely at Tesla’s materials, but agreed that they look precariously low to the ground. Some skirting is readily removable (5-15 minutes). Again I don’t know about Tesla’s.”

          Both of the major changes to the tesla truck to reduce drag have nothing to do with “electric”. So why haven’t current truck manufactures made these changes?

          There are hints, buried in the BS. They’re not practical.

        4. bob

          “but I have taken a test drive in a Tesla Semi,”

          http://cfdtrade.info/2017/11/teslas-semi-truck-perspectives-cab-and-the-truck-stop-dock.html#comment-2892054

          I haven’t had the chance to look closely at Tesla’s materials, but agreed that they look precariously low to the ground. Some skirting is readily removable (5-15 minutes). Again I don’t know about Tesla’s.

          http://cfdtrade.info/2017/11/teslas-semi-truck-perspectives-cab-and-the-truck-stop-dock.html#comment-2892327

          What sort of test drive was it where you couldn’t look at the truck?

    7. doug

      Thank you. Our host knew how the article would come out before he started writing it. Lambert has a history with his T writings.

  9. Bill Carson

    You know what kind of vehicle I think Tesla ought to design? School buses!

    School buses run relatively short routes, shorter hours than city buses, and are parked at the bus barn every night, which would be great for charging. Blue Bird has already unveiled a design for it’s ubiquitous yellow student-mover. Why shouldn’t Tesla dip a toe in that market?

    1. WobblyTelomeres

      You may be onto something. Every lawyer’s nightmare is defending a client who center-punched a school bus full of nuns. Said lawyers will influence system design, hence, school buses may very well be the safest way to travel soon.

    2. nonsense factory

      Perhaps for the European and Japanese market – but given that Betsy DeVos is Secretary of Education, and the lamentable state of public education infrastructure in the United States in general, I can’t see too many orders being placed for school buses in the near future, at least. Since the current U.S. military budget is $611 billion (CNBC), it would make more sense, fiscally, to produce electric military vehicles.

      I’m not sure if that calls for a sarcasm tag or not. . .

    3. doug

      bid purchase by govt, razor thin margins historically.
      others are heavily into this. damlier owns thomas bus in NC, and is putting big $ into electrifying.

  10. YY

    Although not so much when dealing with tractor trailers, I do rely on seeing the movement of the driver’s face/head in their mirror when deciding to pass on the driver’s side in my car or on my bike.

    I don’t know, but would guess that it is very little, how much thought has gone into the reliance upon what would be face to face communications that occur very often between vehicle drivers to signal awareness and negotiate whether or not one has clear passage. This as everyone knows occurs at traffic signal-less intersections on regular basis. One even relies on subtle signals of intermittent usage of turn signals as when one turns it off to give up on a particular lane change until the other vehicle passes. Head movement (eye movement is more problematical as they are not all that visible) of the other driver as well as your own is a communications tool when driving and negotiating traffic. Autonomous vehicles would simply barge along on basis of what it thinks is safe without a clue as to what would be normally very predictable behavior of other vehicles on the road, the drivers of which communicate with head movement and sometimes by nudging their vehicle in a direction that says intent.

    I see myself flailing madly at another car, trying to give them the right of way at a four way stop, not knowing that it is a robot with zero awareness.

    1. ger

      The usual suspects will be bought up in congress and state houses. It will be your fault if you are run over by one of these robots.

      1. Lambert Strether Post author

        > It will be your fault if you are run over by one of these robots.

        Yes, that’s where we’re going to come out. OTOH, none of our young people can afford cars, and many don’t seem to want them, so perhaps this is a problem that will solve itself.

    2. XXYY

      I don’t know, but would guess that it is very little, how much thought has gone into the reliance upon what would be face to face communications that occur very often between vehicle drivers to signal awareness and negotiate whether or not one has clear passage.

      This is one of the few cases I can remember where this aspect of autonomous vehicles has been discussed.

      Bad as it is for car and bicycle drivers, the situation is worse for pedestrians who are interacting with cars. We have all had the situation where we are trying to decide whether to walk in front of a stopped car, and relying on eye with the driver to conclude that it safe to proceed. This happens almost daily. Would anyone on foot care to cross in front of a stopped car if there were no human driver? Or if the driver were obviously heads-down, tweeting on his phone and the vehicle under command of an AI driving system? This extremely common situation gives me the chills.

      Driving is very much a social and human activity, something the autonomous vehicle people can’t seem to get.

  11. nonsense factory

    You’d think a central issue would be the difference in fuel costs versus electricity costs per mile, and that those numbers would be prominent in any discussion of the virtues of electric vs. fossil-fueled semis – it’s such an obvious point, that a lack of discussion is really disappointing.

    Obviously, the fossil fuel industry doesn’t like the idea of the loss of revenue from trucking fuel sales that electric semis would represent; but an analysis of the scale of such losses seems appropriate. A little searching turns up Musk’s claims:

    More significantly for trucking companies, Musk said that a standard diesel truck would be 20% more expensive to operate than a Tesla truck: $1.26 per mile compared to $1.51 per mile.

    Musk said that if Tesla Semis were to drive in a convoy, in which trucks can draft off each other while letting computers and sensors keep them just feet apart at highway speeds, it could compete on cost with trains. “This beats rail,” said Musk, throwing up an 85 cents per mile cost. – USA Today

    So if these claims are incorrect, I’d expect to see a lot of push-back. Does this mean that’s an accurate estimate, or is this just a topic that fossil fuel investors don’t want to talk about?

    Currently diesel accounts for about 21% of the petroleum-based fuel consumed in the United States, so a significant shift to electric semis would seriously impact demand for fossil fuels. Given that so many fossil fuel production strategies – fracking, tar sands, etc. – are near-margin in terms of profitability (i.e. if production costs are on the order of $60 a barrel, greater than current crude oil pricing, they are losing money), any slump in global fossil fuel demand makes a significant fraction of these strategies economic losers.

    This also brings up the topic of how investment funds operate – if they’re investing in fossil fuels, and also in auto and truck manufacturers, then where do they see the balance sheet – i.e. if they choose to invest in electric vehicles, then the smart move is to also get out of their fossil fuel investments – but if they’re heavily into fossil fuels, the last thing they want to see is truck manufacturers shifting over to electric manufacturing. Given that the large shareholders appoint the corporate board members who make such decisions – well, is this really a winning strategy in the long run?

    This is also a strong argument for employees of the corporation being the majority shareholders, so that their company can remain competitive as the world transitions to electric vehicles. Given that China is going to implement bans on fossil fueled-vehicles (for reasons of local air pollution), demand for fossil fuels and fossil-fueled vehicles has nowhere to go but down. This doesn’t bode well for expensive fossil fuel production projects, either.

    1. Jeff N

      Don’t forget that processing a barrel of oil always yields some gasoline and some diesel. What are we gonna do with the unused diesel?

      1. nonsense factory

        Oil refineries have a huge degree of control over what they convert crude oil into, but it does depend on the characteristics of the crude oil – lighter is easier. But this is certainly a problem with the heavier crudes (tar sand, high-sulfur heavy, etc), which already require lots of processing and have huge up-front production costs. The solution then is to stop producing heavy crudes – leave them in the ground.

    2. Anon

      …it could compete on cost with trains.

      Well, trains run on rails that are privately funded, trucks on the Interstate only pay a portion of the cost of the highway. In fact, if the Interstate carried just passenger car traffic the cost of road building would be substantially cheaper. (The depth of road base and concrete/asphalt is a function of the weight and number of vehicles being carried.)

      Putting 40,000 lb. trucks in close convoy is not only dangerous (accidents happen) but puts increased stress on the road surface/base. And convoys are only legal in a few states. (Nevada allows them but only on the Interstate and only in restricted zones.)

      A final point: trains are not powered directly by diesel engines. A modern locomotive is run on electric motors connected directly to the drive-wheels; the electricity comes from a diesel powered electric generator that runs very efficiently between about 75 and 250 RPM. Yeah, they are the original hybrid vehicle.

      1. Lambert Strether Post author

        Yes, always look for the subsidy. The public will end up paying for every single one of the infrastructural improvements needed to get robot cars and robot trucks to work.

    3. LifelongLib

      Seems like a shift to electric vehicles only makes ecological sense if there’s a corresponding shift to renewable/non-polluting ways of generating electricity. Otherwise you’re just burning more fossil fuels to make the additional electricity needed, maybe with less overall efficiency than you’d get by having the vehicles burn the fossil fuels directly…

      1. PlutoniumKun

        Its not quite that simple. as its easier to promote off-peak charging for EVs (and use car batteries for power buffering). Put another way – the more electric cars there are, the easier it is to integrate solar and wind into any given network system. In any event, research in Europe indicates that the promotion of EV’s reduces carbon emissions significantly

    4. Lambert Strether Post author

      > if Tesla Semis were to drive in a convoy, in which trucks can draft off each other while letting computers and sensors keep them just feet apart at highway speeds

      Sounds like “civilians” on the road might have a hard time with this. There’s enough resistance to adding additional trailers already.

      On the economics generally, I’m sure there will be plenty of pushback. As I said in the beginning of this post, I wanted to focus on the Semi as a vehicle. I like assessing the quality of a work product, since (IMNSHO) it’s the best possible indication of the nature of the institution that produced it (at least for a humanities major).

    5. PlutoniumKun

      The potential of ‘convoying’ has been researched for quite some time – it was a central part of thet, arguably the first major study into autonomous vehicles. Its many years ago since I sat in lectures on transport engineering, but I recall that there were very impressive fuel savings reached with convoying, although it was considered at the time that the technology wasn’t established enough to go commercial (this was back in the early 1990’s) and there were of course regulatory obstacles. Its a very obvious potential advantage for partially automonous trucks, especially in areas with big distances and good quality highways.

      This is one (of many) reasons why I think we’ll see semi-autonomous trucks on the road many years before it will ever be viable for cars. The potential economic benefits are very real while the technological obstacles for fixed route deliveries are significantly less than for private cars or Ubers or whatever.

      1. vlade

        There was, a few years back (I believe well before any passenger self-driving cars were allowed on the street), a cross-europe automated convoy experiment. It didn’t raise much news as I remember.. I agree with you that truckies should be the one worried, not taxi drivers – althought the system I heard of as most promising was as you say semi-autonomous, basically a highway-based convoys, where trucks would split off/merge in closer to the local delivery. Sort of not dissimilar to loading the truck(s) on a train ;)

      2. QuarkfromDS9

        PlutoniumKun, I’ve read a lot of your comments about the European Prometheus Project and would love to pick your brain about it and “autonomous” vehicles.

        Don’t suppose there’s anyway to reach you over email or something?

    6. Lynne

      Musk said that if Tesla Semis were to drive in a convoy, in which trucks can draft off each other while letting computers and sensors keep them just feet apart at highway speeds, it could compete on cost with trains. “This beats rail,” said Musk, throwing up an 85 cents per mile cost. – USA Today

      Until one of them hits a deer. Somehow, I doubt his fancy computers can override the laws of physics sufficiently to immediately stop a convoy of trucks driving feet apart at 60+ mph. There are all kinds of safety reasons that hypermiling is illegal.

      1. Mo's Bike Shop

        Imagine being employed to be the ‘driver’ at the front of a ‘segment’ in such a convoy. Sitting for hours and hoping the robots don’t screw up.

  12. Altandmain

    There is one other very big problem that Tesla has.

    They have this far been selling to customers who are very enthusiastic about Tesla cars. These buyers, although small in numbers are very vocal about their car choice and quite willing to overlook the flaws of existing Tesla cars. One quite bluntly confessed to me that fit and finish does not match a car like the top trim of the Chevy Impala, much less a luxury vehicle in the same price category as the Model S or X. The reliability issues have been forgiven, as gave the missed deadlines and the fact that the self driving feature is way below true self driving capabilities.

    This will be a lot more difficult in the B2B market, where the customers are from a very conservative industry. If there are major flaws between the concept and the reality, that could cost sales. Likewise, the long warranty that Tesla is promising – there will be a lot higher standards in the trucking industry because downtime is expensive (simply put if the truck is not on the road, it is not making money and is actually costing money in repairs, along with depreciation costs).

    This is also true of the Model 3. The lower you go down the price chain, the less forgiving the general public is. If the Model 3 wants mainstream appeal, then it will have to have rock solid reliability. Compounding the issue is that Tesla right now does not have that much cash to try to get the car to market and to try to implement a truck which would be another very expensive proposition.

  13. Alfred

    ‘“Anime and ‘Battlestar Galactica.’” Ouch!’ That is a key comment. The image shows not so much a vehicle as a cartoon version of a vehicle. Measured against the basic principles of functional product design, one finds all of them violated and that as a result this ‘product’ is not a product at all. It’s merely an image. Its aesthetics, which emphasize curvature, are remarkably similar to those that dominated the digital-animated classic, Toy Story. Like Toy Story, the Tesla Semi-Truck has a retro look, harking back to those early horseless-carriages of ca. 1900 steered by a centrally position rudder but at the same time the Depression-era visions of Norman Bel Geddes. Hence contemplating it we get a whiff of Back to the Future, another filmed fantasy. The semi-truck’s closest cultural equivalent may be the scenographic ‘house’ in the film, Mon Oncle. Its not merely retro but actually reactionary politics, meanwhile, are revealed by the centralized ‘control’ position to be absolutist (the historical reference being Versailles, where all of nature and all of culture converged upon the singular person of the Sun King).

  14. some lurker

    If they have developed cars that can park themselves (with sensors, LIDAR, etc) what makes people think this truck couldn’t back itself up with minimal help from a human driver? As to the blind spots, why assume the cab and trailer aren’t bristling with collision avoidance sensors that would allow a driver to know what’s on either side of a truck at the same time? Can your truck do that? You know that most drivers ed courses tell you not to pass a truck on the right? So much for the excellent situational awareness and lack of blind spots.

    I concur with the poster above that tech criticism might not be in NC’s wheelhouse: critique Tesla as a business but leave the products alone.

    1. rusti

      what makes people think this truck couldn’t back itself up with minimal help from a human driver?

      There are about three semi-trailers for every semi-tractor, so unless you’re hooked up to a Tesla-specific trailer you will be limited to the bottleneck of the standardized CAN communication link between the tractor and trailer barring an extraordinarily expensive retrofit for every trailer you plan on hauling (to add additional sensing elements like cameras or LIDARs together with a viable communication link). I’ve worked with some function developers for automated or partially automated backing functions, and it’s not an easy problem because even things like the articulation angle between the tractor and trailer require new sensing elements and don’t come “for free” and worst of all you’re running blind on the trailer unless you make it easier for yourself by equipping it with expensive hardware.

      Often it’s the case that the tractor operator doesn’t own the trailer. It could be done for a test trailer, but then it would just give even more credence to Hubert Horan’s criticism that the underlying economics are abysmal because someone would have to add all this junk to any beat-up old trailer you’d want to haul around to use the functions.

    2. Lambert Strether Post author

      > If they have developed cars that can park themselves (with sensors, LIDAR, etc) what makes people think this truck couldn’t back itself up with minimal help from a human driver? As to the blind spots, why assume the cab and trailer aren’t bristling with collision avoidance sensors that would allow a driver to know what’s on either side of a truck at the same time? Can your truck do that? You know that most drivers ed courses tell you not to pass a truck on the right? So much for the excellent situational awareness and lack of blind spots.

      Does this help? (I eliminated the handwaving to expose the essential logic.)

  15. bob

    “…but the cameras face backward….they don’t get mud/salt/rain on them”

    Then why do ALL SUV’s all have back window wipers?

    What happens in the broad space between reality and theory seems to be elon’s downfall. That’s before putting a price tag on it.

  16. ChrisAtRU

    Please permit me a moment of dystopian tongue-in-cheek … ;-)

    Once you understand/accept/believe that Tesla isn’t positioning itself for a future where its semi product needs “drivers”, the departure from actual human ergonomics is not surprising here. Musk and his ilk are all about “robots” replacing humans – although Tesla won’t fully succeed in that endeavor here … yet. The center seat is really designed for the next generation of passive human assist – there to reboot/triage/recharge/protect-the-package, but not actually manage the motor function as it were. Non unionized, semi-stowaways working for less than a living wage without benefits should fit the bill. It’ll be touted as easy work:
    – Get to travel and America’s far flung interstate destinations!
    – Comfortable spacious interior!
    – No background or credit-check required!
    Training for the job will be two days:
    – One to prove you can learn and execute the trouble-shooting scripts
    – And another for signing forms, waivers and peeing in a cup

    DISRUPTIVE!

    #TheFutureIsShite

    1. Lambert Strether Post author

      Gee, that’s odd. Elon’s not marketing the Semi that way!

      I guess a detailed examination of the Semi as an artifact does pay off, doesn’t it? No wonder I keep getting told not to do it; “not in my wheelhouse,” etc.

      1. Mark P.

        Lambert wrote: Elon’s not marketing the Semi that way!

        Come on, Lambert. However Musk markets his Semi (should he actually bring it to fruition), clearly any human on board an autonomous truck will be there in a supernumerary role as a watchman/guard. That’s the point of the technology.

        More generally, because you don’t like the idea of AVs, you’re clearly trying to find objections to them. But because it’s 2017 and you’re an older guy who doesn’t go to AI trade shows, you imagine that the technology is far less advanced than it actually is. As a result, you’re asking the wrong questions and coming up with the wrong answers about where the problems are/will be. Let’s forget Musk and look at Mercedes Benz’s efforts with AVs —

        [1]
        Here’s a headline from last year:

        This is a reference to the trolley problem, that old Philosophy 101 chestnut, which autonomous cars have turned into a real practical dilemma so that ethicists, AI experts and lawyers nowadays have conferences on
        .

        However, the world’s oldest automaker has just announced it doesn’t view the trolley problem as a problem. To the contrary: Mercedes-Benz headquarters in Stuttgart, Germany, claims that when it starts selling autonomous cars this year, their ‘moral algorithm’ will always prioritize their occupants’ safety over any number of pedestrians’ lives. What’s up with that?

        [2] The CEO of Mercedes-Benz USA, Dietmar Exler, gestured in the general direction of the answer when he claimed that what’s really delaying autonomous vehicles “is humans.” If autonomous cars were deployed today, Exler said, they’d share the road with human-driven cars and humans would “bully” driverless cars (which will be programmed to follow laws and avoid obstructions) by speeding, driving erratically and cutting in line in traffic.

        Exler is being somewhat ingenuous. But basically he’s right. Consider some of the problems, especially from an auto-manufacturer’s POV.

        [3] For starters, one study, ’ found that people favor minimizing number of deaths as a general principle, even if that jeopardizes an autonomous car’s passengers. Yet when asked what autonomous cars they’d personally buy, the same people want one that prioritizes protecting them and their passengers.


        [4]
        This is a classic cooperate-or-defect social problem. Furthermore, in any and all potential scenarios involving AVs versus human drivers, reflexivity will be central to whatever real-world outcomes ensue. (Reflexivity, meaning that each actor’s next action will be predicated on the previous actors’ actions.) In short, we’re in the realm of game theory.


        [5]
        So what’s that mean? Here’s a very simple example: an autonomous vehicle is parked at the curb but preparing to pull out into a busy street. Though its indicators are on, human drivers nevertheless keep coming: the autonomous vehicle, programmed to keep specific safe distances between itself and other cars, can’t move into traffic. (This has actually happened).

        [6] Now let’s reconsider trucks. If all it takes to slow and stop an autonomous vehicle is to drive too close or throw on the brakes ahead of it, large trucks transporting expensive goods will be easily stopped and looted by thieves, and kidnappers will have straight shots at those wealthy one-percenters who comprise the upper end of Mercedes-Benz’s customer base.

        So manufacturers will be in the business of offering different versions of their ‘moral algorithms.’ After all, simple physics tells us that just different vehicles’ weights and braking times will require variations in how their moral algorithms perform. Large trucks, in the same way that they’ll require longer braking times, will be programmed with a wider range of avoidance strategies than just stopping dead when another vehicle gets too close. The risk of criminal attack will be factored in, too.

        [7] But then consider the sorts of questions that arise. If manufacturers offer different versions of their moral algorithm and a buyer knowingly chooses one version over another, is that buyer to blame for any harmful consequences of the algorithm’s decisions? Why would the blame not necessarily attach to a manufacturer that created different, ‘more aggressive’ versions of that algorithm in the first place?

        Because manufacturers are going to compete to offer ‘better’ AV algorithms and AI. Do you begin to see the problems?

        (And yes, the insurance industry is going to factor in to how this all plays out.)

  17. SteveB

    Every time I read an article about autonomous vehicles I have the same thought.

    Does anyone else remember back (in the 70’s) when they instituted the 55 mph national speed limit.
    Three guys drove exactly at 55 mph three abreast going across country on route 80. The traffic back up behind them was monumental. Of course they did it in protest…

    But I see monumental traffic jams all over the country when all these autonomous vehicles obey traffic laws to the letter.

  18. Katsue

    Minor nitpick, but I think in line 2 that you mean laudatory coverage rather than laudable (praiseworthy) coverage.

  19. michael

    Just a couple of thoughts.
    When, if ever, will the infrastructure such as interstate highways be upgraded or expanded. Who’s going to pay when politicians don’t usually drive our congested roads. They fly.
    Not enough money is spent now so how do we expect the government or, god forbid, private companies to build new roads or expand highways to three lanes etc.
    Truckers here along interstate 5 in California continue to pass other trucks and leave a long trail of cars behind them until they finally pass the “slow” truck. The CHP never ticket them.
    Don’t you think the Tesla semis would have a self back up system? Cars parallel park today so why can’t trucks auto back up? They would also have side car and blind spot warning systems installed.
    I’d rather take my chances with an automated computer driven semi than with an over worked, stressed out, low paid(which may be debatable), lane cutting person driving a semi.
    Where will we get all the electricity used on electric vehicles and trucks.
    Automated computer assisted or self driving trucks will be safer for the other drivers on the highways and byways.
    Safe driving all.

  20. tegnost

    “I’d rather take my chances with an automated computer driven semi than with an over worked, stressed out, low paid(which may be debatable), lane cutting person driving a semi.”
    sounds like you have some priors here. Maybe you should tell us your income so we know where to place the ostensilby low paid kajillionaire driving a semi back and forth across the country in a randian self sufficient manner. Funny that in your place as an american consumer you feel wronged when one of the trucks carrying your goods to wal mart, (where you must shop because cheap, right?) gets in the way while you live the consumptive california lifestyle relying on cheap labor and cheap energy. I have driven up and down I-5 countless times in my 57 years and have not experienced the pain that you refer to in having to wait your turn on the interstate. You close with the fantasy that a self driving semi will be safer when today, nov 28 2017, there is no safe self driving vehicle, as in no human, in existence, and numerous fantasy land assessents of what they will be able to do with features that have not yet been invented…f-35 anyone? Why don’t you just say I hate truck drivers and humans in general and can’t wait until anyone I think is stupid is replaced by a robot “…bristling with sensors” as a previous poster put it. And as someone who has navigated siskiyou summit and driven down past ashland in a snowstorm I would much rather have a person in control of that 80,000 lb projectile hurtling will ye or nill ye down that hill. Just arrogance and hubris to claim that you drive great and everybody else is the problem. If you drive in california, you cut people off, impossible not to. Your self driving technology is not as competent as that deplorable behind the wheel (who, by the way, is actually bristling with sensors). You are probably more at risk from mexican trucks being allowed to drive on us highways, which you certainly favor as you are a californian and every californian knows that mexicans are cheaper, and cheap is what you’re after. And since we’re on the topic of cheap, where are these trucking companies with barely legal rigs going to get the money to buy a fleet of tesla semi’s? I’m in construction and every company almost has some real POS unsafe trucks because that’s what they could afford to buy, transfer that reality over to your fantasy and see what it looks like. How about all those tomato trucks in the CVBB, yeah, those guys will all be self driving once archer daniels owns every acre of land in the valley and can scale their self driving fleet globally, but at some point there will be a demand issue (as you pointed out, truck drivers must get paid, and then they buy stuff), but I imagine that can be solved with mandatory purchasing, as it was with obamacare. Having been cheated by numerous cheapskate californians (I’m not rich, bill in the mansion next door is rich), I seem to have some priors myself. All for me and none for thee. As cali goes, so goes the nation, and cali is corrupt to the core (see CalPers, see difi’s husband selling post offices for actual kajillions, not truck driver kajillions, I could easily go on). The tesla semi center console looks to me exactly how my programmer friends look like at their desks (feature?), as a matter of fact, if you’re a programmer and the self driving semi could actually be your office and you could get paid a truck drivers kajillion and a programmers actual kajillion and if you added the two together you would have more kalillions and not have to share the highway or anything else in true california style… And in closing I will implore you, should your self driving semi actually take the road in numbers, please do not cut off the self driving semi just because you can. Wait your turn, it’s safer.

    1. michael

      In response to your comments. In no particular order.
      Well, you and I both pay for the insanely overpriced and under performing F-35. A fantastic waste of money. Why do we spend $600 billion on the military and another $200+ billion on our intelligence and black box agencies?
      Through my respective elected politicians I have voiced my displeasure with the military and intelligence agency budgets and the F-35 in particular.
      I live on Social Security and a pension. I am not rich, don’t have a million dollar beach front home or a mansion on a hilltop. I will have one when I meet my maker for he has many mansions.
      If I live on cheap labor here in California and also from other third world countries I guess you do to.
      I drive 400 miles a couple of times a month to help care for my mother who is suffering from Alzheimer’s disease and my father, a 32 year Army highly decorated combat veteran. Silver star, Bronze star with V for valor and 4 Purple hearts. He doesn’t brag about it but he is a hero to me and is happy to see me arrive home safe.
      I prayed every day for two years that he come home safely. My father has Multiple Myeloma(look it up) caused in part by being exposed to agent orange. Helping them is my duty.
      Cheap labor. How about you, or one of your kids or any friends you have try working the agricultural fields for a few weeks. You won’t last. My father(a Mexican born a US citizen) worked the cotton fields in New Mexico and my mom, along with my bother and me, worked bunching onions,cauliflower and broccoli in the fields of the Salinas valley. Oh, we got paid “piece” rate.
      The field workers should get a fair and living wage. Just like everyone of us who is willing to work.
      I never said I was a great driver or that everyone else is the problem.I have one driving ticket.
      Both the Obama and Trump administrations have allowed your “Mexican truckers” drive in the U.S.
      Looks to me like you don’t like Californians, stereotype all of us as cheap and at the same time rich millionaires, hate Mexican truck drivers and those who work our fields.
      As one decent American once said, “Have you no sense of decency”.

      1. tegnost

        Oh don’t get me wrong the mexicans and I get along great, we’re just workers trying to make it and none of us responsible for the policies enacted which created the current paradigm upon which we gaze in wonder. My issue with mexican trucks is the regulatory arbitrage where maintaining your fleet in mexico may be at a lower standard and also is intended to pressure american wages since as I recall you mentioned overpaid truckers. You’re also correct that I have “issues” regarding California. I still say that self driving technology has significant obstacles that call it’s inevitability into question.

  21. vomkammer

    Doing an 11-hour stint in a dark cockpit in the glow of large digital screens only works in anime and “Battlestar Galactica.”

    Actually, not so different from flying a modern airliner at night.

    It makes sense that Semi Truck cockpits resemble (in design and operation) their airliners counterparts, with the computer does the rutinary work while the human monitors and acts on abnormal conditions.

    So, I do not see why the design principles and human factors considerations that have improved airliner safety during the last decades cannot also do it for Semi Trucks. Of course, with proper adaptation.

  22. Jeremy Grimm

    I know unions aren’t what they used to be — but wasn’t the teamsters union one of the unions often inclined toward physical confrontations with scabs? The new autonomous trucks might need to include autonomous systems for self-defense.

  23. Jeremy Grimm

    The more I think about the more electric trucks — human driven or autonomous — trucking goods all over the country just seems so retro. The near future without cheap petroleum will probably end all that whether the motor runs on electric, hydrogen, natural gas, gasoline or diesel. I suppose if we plant enough fast growing trees we could always run steam engines — if we are still spending our energy and energies on transport to exploit cheap labor from across the oceans and far far away.

  24. Edward E

    Build fast, fix later: Speed hurts quality at Tesla, some workers say

    The luxury cars regularly require fixes before they can leave the factory, according to the workers. Quality checks have routinely revealed defects in more than 90 percent of Model S and Model X vehicles inspected after assembly, these individuals said, citing figures from Tesla’s internal tracking system as recently as October. Some of these people told Reuters of seeing problems as far back as 2012.

    I try but still cannot wrap my head around talking about all this convoluted automated system crap yet. The good ol days of the check call and paper logs, millions miles since with no preventable accident. During a couple of solar storms the electronically powered stuff we already have to use were all malfunction junction until the flare passed. My first stints at trucking were with Pete’s that had flamethrower stacks and going 110mph, it could go faster but that was my speed limit. When certain it wouldn’t attract bears. Going from Fresno to Chickendale only stopping for a nap, pull the curtains back and see the blowing snow, get going again trying to stay ahead of it. What a thrill it was being young!

    Best twuck I ever drove was an 1992 approximately FLD Shaker called the Moho never needed anything but PM’s. Trucks today are junk.

    1. Edward E

      I meant to say that the Moho was built in Canada. It was a REAL pleasure to drive because of being so reliable. You didn’t have to worry about some crap piece of fairing, side skirt or boat tail breaking off and flying like 65mph frisbees, possibly cutting the throat of a poor Harley Davidson rider.

      I agree with this!

      Autonomous trucks. The massive hype that has accompanied autonomous trucks is scaring away potential drivers. “Why go to the trouble of getting a CDL when my job will be automated in a few years?,” he wrote. “And the only thing more boring than driving might be sitting in a truck that drives itself.”

Comments are closed.