Tesla Fleet Racking Up The Miles With & Without Autopilot
Tesla has posted an update on total fleet miles driven, which just hit the 2 billion mark. This is exciting because the update came with a chart that beautifully depicts the trend chart over time and forecasted into the future. Over the next 2 years, the trend spikes up very quickly, as we would expect with production doubling every year, as is the plan for the next few years — resulting in total miles driven tracking along an exponential curve.
Speaking to just how crazy the exponential trend this is on actually is, the fleet has gone from 1 billion miles to 2 billion miles in less than a year! Absolutely astonishing considering it took almost 3 years to rack up the first billion miles. That’s 2 billion gasoline-free miles … awesome.
Tesla owners rock! From 1B to 2B miles in less than a year 👏 pic.twitter.com/ERI4KGToGG
— Tesla (@Tesla) April 8, 2016
Equally impressive, the curve tracks right along with what we would expect as the output from the adoption % S-curves (yellow line below) that we have similarly been tracking for EV growth. The S-curve is one of our favorite graphs here at CleanTechnica because so many current clean technologies are ripe for liftoff on the S-curve trend line, implying disruptive growth. Electric Cars, Solar, Wind … as well as a handful of additional technologies that are on the horizon.
Worth noting, given the driving (or rather, traveling) in these electric automobiles — 47 million of those miles were on autopilot. This is almost more noteworthy, as all miles driven in autopilot (of which I may have accumulated a total of 4) give Tesla desperately needed information about the roads cars drive on, how the system is performing, how often drivers have to intervene, etc.
Tesla has shared that autopilot would be sending this information back up to the cloud, which it would use to dynamically improve the system. In the time autopilot has been live, we have seen autopilot dynamically improving day by day, but each car has to learn the same lessons. When test driving a car back in November, a Tesla rep shared that the functionality to use the information in the cloud from other autopilot drivers in order to improve was not yet functional.
However autopilot pans out, Tesla is arguably leading the pack when it comes to autopilot functionality and has shared farther-reaching goals related to autonomous driving than any other auto manufacturer. That’s not to say that things won’t change, but with Tesla already having 47 million miles worth of data from cars actively using autopilot, it has a head start that’s going to be tough to catch up to. (Editor’s Note: in case you missed these, I recommend the two articles from robotics expert Mike Barnard about why Tesla’s approach to self-driving cars seems superior to Google’s approach and the fact that Tesla has >40 times more miles driven on autopilot — gathering data — than Google has even though Google’s testing has been going on for several years and Tesla’s autopilot-equipped cars just hit the road.)
Over 47M miles driven on Autopilot, the more you drive, the more we'll learnhttps://t.co/REuR12rMHH
— Tesla (@Tesla) April 9, 2016
Have a tip for CleanTechnica? Want to advertise? Want to suggest a guest for our CleanTech Talk podcast? Contact us here.
Our Latest EVObsession Video
CleanTechnica uses affiliate links. See our policy here.
“the fleet has gone from 1 billion miles to 2 billion miles in less than a year!”
Not to rain on anyone’s parade, but the global Toyota fleet does this in less than a day… 😉
Yes but on gas . not as impressive or healthy
Your comment is almost like saying that Exxon makes way more money than any cleantech companies out there. Yes, it might be true today but it’s not the future. In fact, it’s damning the possibility of the future which makes it hard to view in a positive light.
It took Tesla 3 years to get to the first billion (~1000x longer than Toyota), 1 year to get the second billion (365x longer than Toyota)…see where they’re headed?
Toyota isn’t smart enough to make or sell EVs…and will get to sit back and watch its market share erode until it does. It won’t happen overnight…but let’s talk in 5 years and see how things compare.
What is important is the increasing rate of the upward trend.
What will it be in 2020?
It’s exponential. It doubles yearly. This will probably slow down.
It sounds way less impressive when converted to barrels of oil per day. Assuming each Model S and X replaces a 20mpg luxury barge and 19 gallons of gas per barrel, that billion miles in a year is only 7200 barrels of oil per day.
Assuming every Model 3 replaces a 35mpg medium sedan, a full year’s production of 300,000 driven the US average 12,0000 miles per year only eliminates 23,000 barrrels/day of demand.
We need a lot more good, inexpensive EVs to move the climate needle. Faster!
Sorry, 15,000 barrels/day for the 300,000 Model 3s.
Vs. 90 million barrels/day worldwide demand.
The other automakers really need to get moving.
Other manufacturers do need to get moving, but the plan of 500,000 EVs from Tesla per year means that every year Tesla will reduce daily oil demand, which may be offset by population growth and other demand factors, but at some point a tipping point will occur where EV supply can materially reduce oil demand.
Obviously just at the start of the curve, but very potentially exciting times ahead.
But replaced it with electricity made from coal. We need cheap nuclear power. Even Bill Gates said so.
New nuclear runs 13 to 19 cents/kWh. More expensive than average retail electricity.
If someone figures out how to make nuclear cheap then we can discuss whether we want to include on our grids (and deal with nuclear’s problems).
Nuclear is prohibitively expensive due to environmentalists fighting it for decades and causing over-regulation, before some of them realized that it was actually a good choice.
No, Peter. Nuclear is expensive because it costs a lot of money to build a reactor and it takes a long time to build a reactor. During construction the ‘overnight cost’, the cost of construction, can double as interest piles up.
Nuclear prices itself out of the market well before there was any significant opposition from the public. Resistance to nuclear didn’t really amount to much before Three Mile Island melted down in 1979. Take a look at the history of nuclear reactors in the US and how their prices rose over the time period they were being built.
.
My friend has a “12.48 kW” solar system. On a good July month it will average 9.6 kW per day (77% of what the installer claimed). It can charge a Tesla about 90 miles per day but cost $53,000 after all incentives. So by fronting that money, he was able to save $7 per day in utility power – at least in the summer. Pay back time is about 30 years, but the system will be obsolete in 15-20 years, so it will never cover its own costs. I wish the situation were different. I would like a system but I like to do the math, and there is no way to make it work.
He paid 53,000 US dollars for a 12.5kWp system?
I feel sorry for him.
Also 9.6kW per day is not a relevant number here.. this is either 9.6kWh which would mean the system is completely mis-designed or you didn’t understand something.
A 12kWp system is expected to average 3-4 hours of 100% power per day on a back of the envelope estimation.. so on a normal day one would expect about 36-48kWh out of it (at least).
If you have an EV you need ~0.3kWh per km.
For a 100km trip this equates to 30kWh.
A pv installation with 10kWp should be able to do that.. daily (if you can charge from it that is).
That’s $4.24 per watt. Not a terrible number. Especially if he’s in an area where there’s not a lot of installing happening.
Peter, in July one might get 6 hours of usable sunshine. 12.5 kW x 6 = 75 kWh. Kilowatt hours is how electricity is measured.
75 kWh at 0.3 kWh per mile would mean that he should be able to charge over 200 miles per day. Between 10% and 15% of the generated electricity will be lost in the inverter and in the process of charging the batteries.
Now the system could be underperforming. The angle could be wrong and the Sun not hitting it very ‘straight on’, a bunch of the energy would be bounced off the front glass. Ideally panels are mounted so that their angle can be adjusted a few times throughout the year to keep them close to 90 degrees off the midday Sun.
There could be shade problem. Or there could be something wrong with the system.
Payback time. You need more numbers to figure that out. You need to know, basically, how much he’s saving on his electricity and gasoline bills. If he took the federal subsidy when he installed. Thirty years is a very long payback time. I suspect there’s an error in someone’s math.
Obsolete in 15 to 20 years? Not at all. Not at all.
We really don’t know how long solar panels last. Our oldest are now getting close to 40 years old. At age 35 they were taken down and individually tested in the lab. They were still producing over 95% as much electricity as when new.
Panels mounted where they are exposed to a lot of UV light and or a lot of wind/snow loading would probably drop off somewhat faster but less than 0.4% per year. (Still output 92% of new at age 20.)
People make a mistake and think that warranty years = expected lifespan. We sometimes get a 36 month warranty on a new car but we don’t find them laying in heap of junk on their third birthday.
What you need to do is to read up and learn more about solar. Check the Home Power site. Home Power is the great grandparent of end-user solar. They’ve got some good tutorial stuff on their site. Or check the Backwoods Solar site. Also good basic information.
I’ll stick a solar map on the bottom. With it you should get a rough idea of how many ‘solar hours’ you should average where you live.
.
System is from 2013. He averaged 44.4 kWh per day last July.
At no point in July did the system output more than 11.2 kW. At no point in its lifetime did it output more than 12.99.
The panel angle is not completely optimal due to architectural constraints.
How many kWh does it take to charge an 85 kWh Tesla battery?
If the battery were completely discharged (don’t think it can be) then 85 kWh plus ~8.5 kWh for charging loss (10%).
The average US driving day is about 35 miles. At 0.3 kWh per mile it would take about 10.5 kWh per day for most people. It should be possible to generate that much electricity with about 2.5 kW of panels.
Given that he’s seeing outputs over 12 kW at times suggests that the system is working as it should but the overall low performance is due to mounting angle. As the Sun rises and falls (yeah, as the Earth wobbles) throughout the year there’s the one day when the panels are directly facing the Sun and he sees annual peak output.
I used a Kill-A-Watt to check energy input to my Samsung-18650 celled bike battery, and then used a Cycle Analyst to record power out – and it was a 17.2% is lost due to the charger and the battery storage process.
Are you sure that of all the power coming from the wall, that 90% of it will later come from the battery?
And also, MPGe does not seem to account for this loss, which seems like a failure on the part of the MPGe spec.
I see that there are US-based wholesale solar dealers who will sell a 12 kW kit for $20,000. And then I see that you can go right to China and get 10 kW for under $10,000 – the trick being to know how to order the right stuff. I imported a laser cutter and it worked out well and I saved about $10,000. So why not buy something like this, and then hire someone for $5,000 to install it?
http://www.alibaba.com/product-detail/10KW-Price-Of-Solar-Power-System_60265364617.html?spm=a2700.7724838.0.0.LPA6Ut
The problem with purchasing anything from anyone is whether what you pay for is what you get. And if you have a problem how do you solve it?
I purchase stuff from China but it’s small dollar stuff so if the quality is not what I expected then I’m not out much. So far I’ve gotten what I expected. I did have one computer component that wasn’t working correctly and the company shipped me a replacement at no cost.
People would need to do their own due diligence.
—
Sun Electronics and Wholesale Power (both US companies) are selling panels for <$0.75/watt. 10 kW of panels would be less than $7,500.
http://www.ecobusinesslinks.com/surveys/free-solar-panel-price-survey/
(Prices change frequently)
Looks like 10kW inverters run around $3k. Now you've got the big parts of a 10kW system for under $11k. And US companies with reputations you can check.
Five thousand for an installation sounds awfully high. But I've never hired that sort of stuff I. I muddle through on my own….
Thanks for the link.
I was estimating 100 man-hours at $50 an hour but that is just a guess. Seems like my friend paid about $35,000 for installation, so $5,000 is way less.
As for inverters – I figured micro-inverters is the way to go, no? Figure $6500 for those.
Ron put up a video at one time that showed a 5(?) kW system installed on an Australian roof, start to finish, in one hour. (I think that was panels only, not the wiring.)
Roof time, as long as it’s a pretty standard installation should be less than two hours. If the workers know what they are doing. If you’re using racks then the racks get screwed to the rafters. Panels connect to each other with male/female connections. Plug. Done. Panels snap into place on the racks. Shove down. Done.
Some time to run conduit and pull wire. Could be an hour or could be something more if the run is complex. Hook up to the inverter and tie it to the service. An hour of electrician time.
I think microinverters only if you have some sort of a weird shade problem where all panels may not be getting the same amount of sunshine for significant periods.
With today’s panels you can string wire them. Just plug one to the next in series and you end up with a lot of voltage and not a lot of amperage. That lets you run a smaller gauge wire (cheaper and quicker).
It also gets the inverter inside the building rather than out in the elements.
Get some other opinions on microinverters. I’m not a fan.
Are we not talking about 40 or so panels? it would seem like just measuring out how to center them on the roof could take an hour. Just carrying them up a ladder could take another hour. Installing one every 15 minute would be 10 hours. Ok on the single inverter and high voltage.
Shouldn’t take long to do the layout. That job is mostly done by figuring out how you want them racked. Two rows of 20 each? Order the racks. Find the rafters and screw them down.
A job that large and you’d probably use a lift. A lot cheaper and quicker to use a lift than to carry panels up a ladder.
Installation with plug to wire and snap in racks is incredibly fast.
It was doing 30-58 kWh per day in July.
Because the thing we most want is a nuclear plant built with only private oversight, by the lowest cost bidder, and operated by someone who pinky-swears they’ll put away enough money to decommission and rehab the site — starting next quarter…
Everything the government does costs at least three times more than necessary. Look at the F35 program. Even my driveway – government made me put an underground drain system in and turned a $6000 driveway into a $42,000 driveway.
None of the electricity in my area is from coal. ZERO. Last coal plant in New York State closed this month.
For a 2014 roundup of how clean the electric grid is for electric cars, see the Union of Concerned Scientists’ “State of Charge” report.
Here’s a link to the most recent UCS report:
http://www.ucsusa.org/clean-vehicles/electric-vehicles/life-cycle-ev-emissions#.Vw1x9c3er0M
As you see, MY carbon emissions are equivalent to a car which gets 135 mpg… and that was before the last three coal plants shut down. I’d love to see an update.
Are you sure NY state isn’t importing power from other states? It happens in many states.
NY state imports some power from Ontario and Quebec. Both provinces closed their last plant years ago. In Ontario less than 5% power comes from natural gas.
Depends where you live. New England during evening peak right now (9pm 4/12/2016) has only a few MW of coal running.
The natural gas needs to get worked on next…
When Bill said it it must be true.
How did you do that graph