Tesla Semi | Hacker News Reader

Tesla Semi

tesla.com

70 pointsmaximente58 comments

I'm legitimately surprised this product hasn't ended up with interchangeable battery packs. Super charging is "fast" compared to regular electrical charging, but when time is money, it seems like that would be "the" limit that costs sales.

Keep in mind a lot of trucks don't sit idle while cargo is unloaded by hand, in many locations they'll just de-couple, pick up outgoing, and be back out of the gate within 10~15 minutes. With recharge downtime of e.g. 20 minutes, they've managed to place themselves into an even smaller niche than they would be already.

By the way, Semis are a massive source of pollution and have proportionately terrible Mpg while idle or towing light loads. So this is an area ripe for improvement, and this is an asperational product. I appreciate a big name trying if for no other reason than it might spur innovation from traditional trucking companies.

Just because I worry that it isn't viable in the market, doesn't mean I'm not a fan.

I'm not going to pretend to be an expert in electric vehicles, or trucking, but it seems that with a 500 mile range, you're getting pretty much to the limit of what a trucker is legally allowed to drive in the United States each day. So maybe recharge time isn't that big a deal, since the drivers aren't supposed to be on the road anyway.

> you're getting pretty much to the limit of what a trucker is legally allowed to drive in the United States each day

That's why there are multiple drivers.

My brother-in-law is a truck driver. Every day, he drives his semi from Portland, OR to Seattle, WA and back. Once he gets back, another driver gets in and does a route.

A single driver may have a limit, but the truck itself doesn't.

How much weight is in the trailer for that 500 miles range? If the answer is zero tons then it is completely meaninless.

The Tesla site doesn't qualify what 500 miles means. For trucks load-range matters, not empty range.

The video says "500-mile range at GVW [gross vehicle weight, I assume] & highway speed", and Musk explicitly says "that's 500-mile range at maximum weight at highway speed, 60 miles an hour". Roughly 06:36 into the video.

And on the subject of charging, it claims "400 miles of charge in 30 minutes" around 13:15.

Over 4 years or so, batteries that have never been supercharged are at about 90% capacity rather than 80% for frequently fast charged vehicles. Given the rapid decline of batteries in their final months, that could be equaling an extra couple years of usable life.

That's an issue in a car, but a massive deal in a semi where they need much, much larger batteries with much, much larger replacement costs.

The real solution here should have been battery exchange contracts. They make facilities to slow-charge batteries and semi's pay per mile or kw he put on the battery. The would be much better for the environment. It would also enable Tesla decreasing the up front cost by moving some of that cost into charging fees to encourage adoption.

https://www.geotab.com/blog/ev-battery-health/

There are a ton of factors that go into battery degradation, not just speed of charging. Also, it varies dramatically depending on the manufacturer - the 4yr/80-90% doesn't line up with real world data from Tesla: https://twitter.com/elonmusk/status/1270784295733301249?s=19

That was my conclusion a couple of years ago. A truck driver legally has a lot of enforced breaks. There are a max hours they can drive in a day. And a max number of hours per sprint.

Which with electronic log books it's hard to cheat on that. And I'm dubious that truck drivers would mind a slightly slower pace as long as it's uniformly enforced. And revamping truck stops to charge trucks seems like something where all the needed technology already exists.

As I recall, one of the major initial target markets for this is regional distribution. A 300 or 500 mile range gives you a 130-230 mile radius for distribution (dropping a bit for reserve in case you have to take a detour or just for safety). That's not a bad range, at least on the east coast. The central states and some western states will be harder to cover with those ranges, but feasible for large chunks.

The logistics of interchanging a $100,000 - 10,000lbs battery needs to be considered.

Most of the value if the tractor is the battery as is the weight, you aren't just going to swap them out at some random truck stop.

300-500 mile range is cringe-worthy when a dual tank diesel will get you near 2K miles. (And yes, truckers aren't able to do this, but filling up is time).

Edit: I'm in the industry and this is one case where time really is money.

How does the math change if diesel trucks have to pay for their CO2 emissions?

Yeah, it would be so nice to have market forces figuring this stuff out. Haphazard subsidies are better than nothing, but I suspect much less than optimal.

Truck drivers have to take mandatory breaks every so often. Seems like after driving for 300 - 500 miles, would be a perfect time to recharge while you recharge.

As others have pointed out, the driver has to rest. The truck does not. It's an expensive piece of capital equipment that you want to keep on the road at all times until it needs maintenance. This is a good example of where swappable battery packs would actually be useful. The extra weight cost of making it swappable is much smaller as a percentage than it would be for a car, and shipping companies that run regular routes could easily have their own facilities to handle such things.

In Norway there is a mandatory 45 minute break after 4.5 hours of driving, and you cannot drive for more than 90 hours during a 14 day period.

Note, the real "killer app" for electric trucks is not long haul, not yet. Diesel is really, really good at long haul. And arguably, since the engines are running at a pretty constant speed for a long time, they're pretty efficient at it.

Electric trucks are first aiming at local and drayage routes, which actually make up a significant fraction of the trucking industry. You don't have to capture 100% of the market to make a huge impact.

Drayage trucks are typically back at their home base every night, and spend their days running containers to and from ports, intermodal yards, and other shipping depots. These routes aren't actually well-suited to the way diesel trucks and conventional transmissions work, with lots of local roads, starting and stopping, idling, etc. They're ideally suited to electrification, even with fairly small batteries. A typical drayage truck may only put on 200 miles in a day, with many even less. And they're usually parked overnight, so recharging isn't that big a deal. (Yes the chargers are still enormous, but they have hours to spend, it's not like you need to charge the thing in 15 minutes.)

Note that pretty much every "traditional" truck maker has an electric program in progress with test vehicles on the road, some more serious than others. Tesla's entry is interesting, but brand loyalty and history may mean it's not a slam-dunk, if the others can deliver something that works well and still "feels like" the trucks their owners know and love. Others may fly the "futuristic" flag with pride. Time will tell!

I know folks have complained about the term Autopilot. It looks like they're pretty explicit about the driver remaining alert.

Enhanced Autopilot helps avoid collisions, a centered driver position provides maximum visibility and control, and a low center of gravity offers rollover protection. Every driver is responsible for remaining alert and active when using Autopilot, and must be prepared to take action at any time.

Right just as regular Teslas... there will be some massive highway accidents from drivers falling asleep at the Semi's wheel expecting auto pilot to take care of things...

This already happens regularly with truck drivers falling asleep, without Autopilot.

But with an autopilot, it seems more likely that you can get away with it.

I don't know why so many people can't understand this simple thing: people adjust their behavior to perceived risk. Things that appear riskier cause people to be more careful, and ones that appear safer make people more likely to be negligent. The gap between perceived risk and real risk is more dangerous than the real risk.

Nevermind that even an aircraft autopilot doesn't remove responsibility for the aircraft from the pilot. Most small aircraft have autopilots which only keep the wings level. An aircraft autopilot could fly you straight into the ground while working 100% correctly.

Complaints are by people who have inaccurate assumptions of the term autopilot. Autopilot originated in aviation, and has always been an assistant to the pilot, never a replacement. The term is used accurately, and complainers are just ill informed. Autopilot has never been what a layman assumed.

> Autopilot has never been what a layman assumed.

This is actually my biggest gripe; it's an awful misnomer that most laymen don't understand.

I really hope they'll rename it, people have died because of the attitudes towards Autopilot and it would help address that.

Edit: they did adjust the wording on the Autopilot page to be a lot more clear.

In February 2019 https://tesla.com/autopilot opened with:

Full Self-Driving Hardware on All Cars

All Tesla vehicles produced in our factory, including Model 3, have the hardware needed for full self-driving capability at a safety level substantially greater than that of a human driver

http://web.archive.org/web/20190219105618/https://www.tesla....

Since then it has opened with:

Future of Driving

All new Tesla cars come standard with advanced hardware capable of providing Autopilot features today, and full self-driving capabilities in the future—through software updates designed to improve functionality over time.

Which helps make the distinction between Autopilot and FSD.

Aviation requires years of training plus training for the specific type of aircraft. Commercial pilots need thousands of hours as co-pilots before they can be in command. Pilots are also in constant contact with coordinators (ATC) or follow exact routes, dozens or even hundreds of miles apart from other aircraft, where contact is not possible (e.g. over the North Pole).

How is this comparable to the relative mayhem on public roads?

Tesla is not selling these trucks to pilots.

I think this isn't a true Highway truck (Long Haul). It would make a great day truck and could be recharged every night. It could still be a regional truck. A company like Walmart that has regional distribution will be the most interested, since they tend to stick around in <500 mile radius. If you're long haul it is likely you will be running diesel.

There is also the fact to consider that truck stops aren't setup for charging. It would require a massive investment. Still very exciting and I can only imagine the range will increase over time.

I know a lot of national gas chains in Canada (petro-can, Irving) have put in electric chargers along the Trans Canada. They'll be a lagging source, but I expect they'll add charging infrastructure for the long haul truckers (while regional eg Walmart, grocery stores, are more likely to charge at the destinations)

I still want to see a system with batteries in the trailer and chargers at the loading docks. Why stop to charge the cab when your next payload could have all the energy needed for the trip? Imagine a company that constantly needs to exchange goods between two locations. With three trailers, one rig, and two chargers, you could have the rig moving almost constantly. Both endpoints would have the duration of the round trip to load/unload and charge.

Those would be very expensive trailers and the current shipping container ecosystem (and semi trailer ecosystem) could not cope with $100,000 unloaded trailers just sitting in unsecured areas like normal trailers do today.

You're right that it would be convenient for the driver, but I don't think it would be convenient for anyone else.

Actual trucking professionals please correct me if I'm wrong about any of that.

I'm sorry but to me it seems to me that a lot of freight miles could be done with rail, of course you can't replace road freight (yet?), but in terms of ton-distances, rail should still be a viable alternative to reduce carbon emissions on large distances, in many cases. The logistics of course are a little hairy, but the carbon situation makes it necessary anyways.

Did some consulting for a railroad ages ago. They way it was explained to me, two of the (many) challenges are:

- the logistics of coupling long trains reduces their flexibility, so favors bulk goods

- trucking gets all sorts of direct and indirect subsidies that the railroads no longer get

I've since wondered if different engines (perhaps EV), different couplings, or some combo could change the logistics calculus.

I remember when the specs were initially announced two years ago. All these automotive researchers published estimates on what they thought the possible range of performance and range could be. Those estimates got absolutely blown out of the water.

Interesting that the disclaimer is quite explicit here:

> Every driver is responsible for remaining alert and active when using Autopilot, and must be prepared to take action at any time.

So don’t call it ‘Autopilot’.

It’s like calling something’food’ with the disclaimer: not to be eaten.

Tesla wasn't even the first one who picked Autopilot name. Chrysler did in 1958. [0]

I'm not so sure why so many people are hung up on Autopilot name. After all Apple's MagicMouse can't do any tricks, and there is nothing intelligent about Microsoft's IntelliSense. It's also already mentioned below that autopilot is prominent on airplanes despite requiring not only one, but two pilots.

[0] - https://www.curbsideclassic.com/blog/history/automotive-hist...

Yes but I don't think anyone expected autopilot to fully drive the car for you in 1958. Now lots of people think fully driverless cars are already possible, and Tesla still markets their cars as having "full self driving" (distinct from "autopilot", but how many people are going to understand that nuance?)

The term Autopilot isn't misleading. What is misleading is that Tesla claims that all cars since 2016 have full self driving hardware and that they will enable full self driving via over the air updates. Basically they have created the impression that FSD is around the corner and this has probably fooled some people into believing that Autopilot is working much better than the disclaimers say. Some people might think standard features like emergency breaking are included in Autopilot and then they crash into stationary objects that Autopilot didn't detect. If Tesla cars actually had "FSD hardware" then these accidents wouldn't have happened.

https://www.tesla.com/blog/all-tesla-cars-being-produced-now...

https://www.theverge.com/2019/1/30/18204427/tesla-autopilot-...

https://www.theverge.com/2018/10/20/18000884/tesla-full-self...

"Autopilot" is also used in planes, despite the fact that we require a pilot and co-pilot to be always active and alert.

Autopilot is a perfect name for it. It matches the original usage of the word in aviation perfectly.

The problem is that people don't understand that even in aviation, autopilot still requires the pilot to pay attention.

Ok, so the name gives the correct implication to pilots and everyone else will be confused, and that's the perfect name for it?

I don't have a problem with the name. If increasing the safety of the system is your priority driver focus detection seems a much better complaint.

But if I have to read another comment explaining that it's fine because it fits the technical definition from aviation I'm going to scream.

It's such a simplistic summation of a genuinely complicated question.

For instance the AP style guide has given guidelines on language to use for partially autonomous vehicles. They're concerned enough to take it seriously.

https://www.caranddriver.com/features/columns/a32702153/the-...

Everyone worried about the charge time/cruise range, I'm just excited for the improvement to urban environments when electric trucks with regenerative braking are the only ones allowed in cities. Diesel trucks (and buses) are too loud for dense urban environments.

Telemetry in trucks is its own cottage industry due to regulatory reasons. With the amount of tech in a Tesla Semi they largely eliminate the need for a separate telemetry vendor. This is an additional revenue stream for Tesla.

Some people are complaining the range is not nearly what is needed for long-range. But it is fine for local and regional trucking, which is a huge source of carbon and air pollution.

Beyond that, batteries are getting better every year, so the range is going to keep increasing until it is long-range.

You know, I think a lot of the people who are critical of the Tesla semi are really members of the fossil fuels forever club who think diesel trucks are just great.

Whats the charge time? And what is the range of a typical semi these days? Is 500mi average?

"Range for a diesel semi truck is limited more by the allowable driving hours for the driver than by the truck itself. It’s not unusual for a truck to have dual 150 gallon tanks. 300 gallons at an average of 7 MPG gives a non-stop range of more than 2000 miles."

https://www.quora.com/What%E2%80%99s-the-average-range-for-a....

Wonder if the legal hour limit will be extended if you're using autopilot

I don't think so since the driver still needs to be alert.

With a 1 MWh battery and assuming a 1MW charger (reasonable given the name "Megacharger" and that 250kW superchargers exists today) should be around an hour. There was some speculation that a [Megapack](https://www.tesla.com/megapack) could back a Megacharger to provide simplified deployment of these chargers to customers.

Quite a bit farther - I think 240gal tanks, 5mpg is fairly common. Costwise the tractor itself seems comparable, I wonder if the plan is to use the Superchargers, or to build charging capacity at truckstops (or both).

Probably both.

I haven't seen all of the Supercharger stations, but the ones I have seen aren't in places where a tractor trailer could park.

The video on the page claims 400 miles of charge in 30 minutes (13:15 into the video).

Is there any more detail on the weight of the battery? I skimmed through and couldn't find anything. If the battery weighs 10,000 pounds, that's a significant chunk of the 80k lb weight they use as a reference

Originally slated for release in 2019, now hyped out of fear of Nikola. Just a few days ago the Model Y was the top priority, now it’s the semi.

Can anyone share details on what the two-year payback period looks like, from the real world perspective of a freight operator? How much money is spent on diesel for a long haul truck freight operation? Obviously it depends on distance. But I have heard that a figure of 8 miles per gallon would not be wildly off the mark.

[2017]

How does it make turns?

Looks absolutely awesome!

We love you, Elon!!!

Reply on news.ycombinator.com