Tesla Semi

[KarenRei]

Do most truckers pull non-stop 11 hour runs, without a stop for bathroom or food? If so, that seems dangerous, and perhaps being forced to stop isn't a bad thing.

In the EU that would be illegal. The EU mandates a minimum of 45 minutes of breaks for every 4 1/2 hours driving. If Tesla can make it so that its semi can keep running indefinitely with 45 minutes of charging broken into one or two stops for that 4 1/2 hours of highway driving, then they have an immense advantage. Fueling costs are around 40% of the cost of running a rig. Amortizing the cab is under 20%. Then there's maintenance and everything else that they have the potential to save. If they can halve fueling costs without affecting throughput in the EU they could literally charge twice as much for their trucks and still make great sales.

I assume Tesla will start (like they do with everything) on the high end - a decked out super-comfortable sleeper cab with tons of power at the wheels. I bet it's going to be gorgeous inside.

Even in the US, without the rest requirements and much cheaper fuel, I can still see good sales. Small slowdowns to stop for supercharging would be acceptable because fueling costs so much (even at US prices); you could funnel the savings into additional trucks and drivers to make up for any lost time.

 

[RubberToe]

Electrek is reporting that Tesla has applied for permission to run platoons and/or autonomous semis in both Nevada and California:

Tesla seeks to test electric truck prototypes for platooning and self-driving in Nevada and California

 

[RobStark]

In the EU that would be illegal. The EU mandates a minimum of 45 minutes of breaks for every 4 1/2 hours driving.

New Hours-of-Service Safety Regulations to Reduce Truck Driver Fatigue Begin Today

FMCSA's new hours-of-service final rule:

• Limits the maximum average work week for truck drivers to 70 hours, a decrease from the current maximum of 82 hours;
• Allows truck drivers who reach the maximum 70 hours of driving within a week to resume if they rest for 34 consecutive hours, including at least two nights when their body clock demands sleep the most - from 1-5 a.m., and;
• Requires truck drivers to take a 30-minute break during the first eight hours of a shift.

The final rule retains the current 11-hour daily driving limit and 14-hour work day.

 

[MP3Mike]

Electrek is reporting that Tesla has applied for permission to run platoons and/or autonomous semis in both Nevada and California:

No, it reported that they have inquired about applying for permission. They have not applied yet.

 

[RubberToe]

No, it reported that they have inquired about applying for permission. They have not applied yet.

You are correct:

Reuters reports that Tesla has met with the Nevada DMV, but it has yet to apply for a license and they are also meeting with the California DMV today.

 

[neroden]

Platooning is illegal for a reason -- it's extremely dangerous. If there's a mechanical failure on the front truck, boom, multi-truck pileup, and they jackknife their way across the entire road and basically shut it down completely for hours.

Thankfully platooning is not necessary to make electric semis compelling. I have no idea why so many technofantasists want to do platooning. Maybe they just have no understanding of safety factors whatsoever.

 

[SPadival]

Research paper of the guy quoted in the Reuters article.

http://pubs.acs.org/doi/10.1021/acsenergylett.7b00432?ref=aelccpJulyArticle

Lithium ion battery researcher Venkat Viswanathan of Carnegie Mellon University said electric long-haul trucking is not economically feasible yet.

“Your cargo essentially becomes the battery,” Viswanathan said of the massive batteries that would be needed to make range competitive with diesel.

When the rubber hits the road in September (during the Semi unveil), I don't know what excuse they will come up with.

I don't have any fancy equations to show this - but If a Model 3 can weigh just a shade over a comparable ICE car, I don't see why they can't scale it up for a semi.

 

[GoTslaGo]

Research paper of the guy quoted in the Reuters article.

http://pubs.acs.org/doi/10.1021/acsenergylett.7b00432?ref=aelccpJulyArticle



When the rubber hits the road in September (during the Semi unveil), I don't know what excuse they will come up with.

I don't have any fancy equations to show this - but If a Model 3 can weigh just a shade over a comparable ICE car, I don't see why they can't scale it up for a semi.

Couldn't resist:

Bumblebees Can't Fly

(yes it's an article saying that it's a myth that scientists proved that bumblebees can't fly, but when I read the blurb about the paper, that's what popped in my head )

 

[neroden]

Didn't we sit down and work out how much space and weight would be needed to move a semi 600 to 1000 miles? It fit in the tractor easily without causing any problem for the axle load, unless it was one of those semis hauling super-heavy cargo.

So, for ordinary goods containers: no problem. For hauling steel beams, maybe an issue.

 

[SwTslaGrl]

IMHO, I think the discussion should move away from generic semis.

For example a fleet owner goes a route A to B, the same year after year and needs a semi with x capability, not more. It's the most cost effective solution.

 

[BioSehnsucht]

I'd think platooning being illegal has more to do with human failure than mechanical, though certainly I can see an argument made for mechanical as well.

At least in the US, it's not uncommon to see two semi trailers pulled by a single tractor. Putting more than two in a line would likely require platooning simply so that they can be maneuvered separately when needed (i.e. all times they aren't going in a straight line), otherwise things are going behave like a train without tracks...

With computer control, there's no reason (other than accounting for all the vehicles around you that aren't in sync with you) why you couldn't drive in formation safely, as long as all problems can be avoided by braking as a platoon as well. The hard part might be simply ensuring that other idiots on the road don't take out a middle vehicle in the platoon ...

 

[kenliles]

Electric truck to debut this year amid boom in EV investments

 

[Curt Renz]

Tesla likes to cut out the middleman. First they'll ferry their own car parts from their Nevada Gigafactory to their Fremont factory, and on the return trip serve as trucking company for goods from Bay Area companies to the Reno Area. Then become their own national trucking company. Due to the low expenses of operating an electric truck, especially when a human driver is no longer needed, they may push most of the others out of business. Walmart operates its own truck fleet, so can Tesla.

 

[Curt Renz]

CNBC - this morning: Tesla meeting with California DMV to test semi on streets

 

[ggr]

I'd think platooning being illegal has more to do with human failure than mechanical, though certainly I can see an argument made for mechanical as well.

At least in the US, it's not uncommon to see two semi trailers pulled by a single tractor. Putting more than two in a line would likely require platooning simply so that they can be maneuvered separately when needed (i.e. all times they aren't going in a straight line), otherwise things are going behave like a train without tracks...

With computer control, there's no reason (other than accounting for all the vehicles around you that aren't in sync with you) why you couldn't drive in formation safely, as long as all problems can be avoided by braking as a platoon as well. The hard part might be simply ensuring that other idiots on the road don't take out a middle vehicle in the platoon ...

australia road train - Google Search:
Such longer road trains are very common in Australia.

Here in the US, I've driven along I-5 and had to go past 4-5 semis in a row with no space between them. In this case human drivers are a much bigger worry than automated ones would be. Also, the humans occasionally get impatient being stuck at 64mph, and pull out to overtake the lead truck at 65mph. :-(

 

[jhm]

Platooning is illegal for a reason -- it's extremely dangerous. If there's a mechanical failure on the front truck, boom, multi-truck pileup, and they jackknife their way across the entire road and basically shut it down completely for hours.

Thankfully platooning is not necessary to make electric semis compelling. I have no idea why so many technofantasists want to do platooning. Maybe they just have no understanding of safety factors whatsoever.

I suspect that the notion of platooning that Tesla is pursuing is as way to put autonomous vehicles on the road. The separation between rigs can be just as much as is needed for safety. The point is not drafting, but piloting. The lead rig has a human driver who oversees the operation of all rigs following it. So the trailing rigs only need enough autonomy to follow safely in the platoon formation.

Notice if the lead rig must emergency brake or take other evasive action, this is communicated immediately to all trailing rigs. It's very different with a platoon of human drivers that can only see one or two rigs ahead and suffer slow human response times. In Tesla platoon, the rear truck will respond just a quickly as the front truck. The latency is measured in milliseconds.

Another way to look at this is to suppose that you already have autonomous trucks. These trucks must still follow a safe distance behind vehicles in front of them. So let's assume that these trucks know how to do this safely. Given that level of autonomy, is there any argument against these autonomous trucks being in communication with other vehicles on the highway? It seems that a communication link only improves safety. Thus, it is advantageous for Tesla to pioneer how these vehicles can communicate with each other. This is informational platooning.

Honestly, I would love it if my Model S could be in communication with other Teslas on the road to improve performance of autopilot. I'm not buying into the hype about how autonomy will enable more vehicles to fit on the road at once. I'm simply interested in how connected autonomy can enhance safety for all vehicles on the road. So Tesla can do a lot of testing along this front.

 

[jhm]

IMHO, I think the discussion should move away from generic semis.

For example a fleet owner goes a route A to B, the same year after year and needs a semi with x capability, not more. It's the most cost effective solution.

Indeed, so much of the discussion around electric trucks is on the extremes of capability, while the bulk of money is to be made on routes that are not so demanding. Fremont, CA, to Sparks, NV is just 248 miles. So a pack capable of 300 miles may be all that is needed for this route. A pack with 600 mile range would actually be suboptimal.

We should always bear in mind the lifetime range of a battery pack. Consider a pack with 600 mile range that has 5000 cycle life. This implies that the life time range of the pack is about 3 million miles. How do you make a 3M mile pack a good investment? If you are using just 300 miles per day, it will take 10k days or 27.4 years to get the full life value out of the battery. It's just stupid to tie up capital for 27 years so you can under utilize it so badly. Even using it 600 miles per day still takes 13.7 years to consume the full life range. This is still really capital intensive. If you were using it 1200 miles per day, it would be consumed in 6.85 years. This level of utilization would have really good ROI.

Note that the argument above really has little to do with the price of batteries. Whether the marginal price mile of range is $80/mile or $800/mile, a business does not want to tie up capital in excess capacity that can't be utilized for decades. Moreover, the price of batteries will continue to come down. So even if you thought you could recover that capital tied up in excess capacity through residual value, you are still competing old expensive batteries with newer cheaper batteries.

So while there may be a lot of manly posturing about how truckers need a 600 mile range or more, the more savvy business move is to buy a pack small enough that you can cycle it about two or three times per day. For example Tesla can supply the Fremont-Sparks route with 300 mile packs. Trucks can make the round trip about twice a day and consume the battery in about 4 years. They can maximize the ROI by speeding up the rate of consumption.

 

[neroden]

I'd think platooning being illegal has more to do with human failure than mechanical, though certainly I can see an argument made for mechanical as well.

At least in the US, it's not uncommon to see two semi trailers pulled by a single tractor. Putting more than two in a line would likely require platooning simply so that they can be maneuvered separately when needed (i.e. all times they aren't going in a straight line), otherwise things are going behave like a train without tracks...

With computer control, there's no reason (other than accounting for all the vehicles around you that aren't in sync with you) why you couldn't drive in formation safely, as long as all problems can be avoided by braking as a platoon as well. The hard part might be simply ensuring that other idiots on the road don't take out a middle vehicle in the platoon ...

OK. So, you basically have two safe choices:
(1) Each truck follows at a safe following distance, where they can stop clear of the leading truck if, say, a bridge falls on it -- or its wireless signal drops out.
(2) Mechanical coupling. It's not widely known, but the mechanical coupling on trains is not *just* to pull them forward -- it's also designed, along with a number of structural features, to prevent them from running into each other. There is also an electrical coupling and an air coupling, the latter of which is designed to send "brake" messages throughout the train. It's fail-safe: if it disconnects, the train stops.

#1 can be called platooning. It gains zero aerodynamic benefits, however.
#2 is never called "platooning".

These are the only safe options because of potential catastrophic failures in the leading truck, *including a failure of wireless communications*, which is a very common thing. Computer control is great and all but if you're relying on wireless, you need to prepare for the wireless communication to go out. There are too many ways that can happen.

Given this, I don't see the point of platooning. Either you're following too close, which is unsafe ... or you're not gaining any aerodynamic benefit from the platooning, and you might just as well have individual autonomous trucks.

(Edit: Jhm suggests a reason. A good one. See next comment.)

 

[neroden]

I suspect that the notion of platooning that Tesla is pursuing is as way to put autonomous vehicles on the road. The separation between rigs can be just as much as is needed for safety. The point is not drafting, but piloting. The lead rig has a human driver who oversees the operation of all rigs following it. So the trailing rigs only need enough autonomy to follow safely in the platoon formation.

OK... I expect cars will zip in between the lead rig and the trailing rig, because Bad Drivers Are A Thing :sigh:

But maybe this is more of a way of getting into the market.... let me follow up on this...

Another way to look at this is to suppose that you already have autonomous trucks. These trucks must still follow a safe distance behind vehicles in front of them. So let's assume that these trucks know how to do this safely. Given that level of autonomy, is there any argument against these autonomous trucks being in communication with other vehicles on the highway? It seems that a communication link only improves safety. Thus, it is advantageous for Tesla to pioneer how these vehicles can communicate with each other. This is informational platooning.

So suppose we have working individual autonomous trucks, but people are nervous about totally unattended trucks. So Tesla says "Well, we'll have a guy in the first truck overseeing all the other trucks." Maybe this makes it easier for people/regulators to accept it?

And then when cars zip in between the trucks in the platoon and they get separated... well, they still work fine. The person in the front is really just telling them where to go, and they eventually meet up at the next truck stop.

I could see this...

Honestly, I would love it if my Model S could be in communication with other Teslas on the road to improve performance of autopilot. I'm not buying into the hype about how autonomy will enable more vehicles to fit on the road at once. I'm simply interested in how connected autonomy can enhance safety for all vehicles on the road. So Tesla can do a lot of testing along this front.

This is a smart idea.

 

[jhm]

OK... I expect cars will zip in between the lead rig and the trailing rig, because Bad Drivers Are A Thing :sigh:

But maybe this is more of a way of getting into the market.... let me follow up on this...



So suppose we have working individual autonomous trucks, but people are nervous about totally unattended trucks. So Tesla says "Well, we'll have a guy in the first truck overseeing all the other trucks." Maybe this makes it easier for people/regulators to accept it?

And then when cars zip in between the trucks in the platoon and they get separated... well, they still work fine. The person in the front is really just telling them where to go, and they eventually meet up at the next truck stop.

I could see this...



This is a smart idea.

It would be nice if Musk would clarify what platooning concept Tesla is striving for. A lot of people are worried about traffic formations that are very inconvenient for other drivers. "I don't want to have to pass ten trucks in a row. It's hard enough just passing one at a time." So there is risk a PR risk, if Tesla does not clarify how their vision for platooning actually makes traffic flow better for all other motorists.

The vision can be radically different. For example if connected cars and trucks communicate well, it may get much easier and safer to change lanes. Your car communicates an intent to change lanes, and the connected vehicles around you immediately open up space for it. These vehicles are not motivated by an ego to block you from cutting in. Even before a large number of cars have this capability, platooning Tesla trucks could be programmed to watch for turn signals of vehicles wishing to cut in, and the platoon graciously opens up space. If a signaling car is in the blind spot of one truck, the truck behind it sees it all clearly and communicates this forward so that all trucks in the vicinity respond appropriately. So there is a lot of potential here for improving traffic flow. It basically comes down to what Tesla intends to program things. Tesla Semis could be programmed to be the most polite, responsive and safest trucks on the road, and that would speak well for the brand.