Tesla Model 3

Status
Not open for further replies.
Hillhater said:
..this is a chicken/egg senario where manufacturers are reluctant to release volumes of new technology (EV’s) into regions that do not yet have the support infrastructure to deal with any problems.
Of course. This is mainly charging location access. The maintenance part isn't as big a deal.
At the same time , Dealerships ( the conventional, ..independantly owned model,.. still used by most manufacturers), are reluctant to invest in retraining technicians or new facilities and equipment, to provide the support for EV’s before there is sufficient demand to justify the investment.
They've already invested. While I had a Leaf, the Nissan dealership was quite capable of taking care of maintenance. There will, of course, always be more people more comfortable taking care of older vehicles because they've been doing it for longer.
It may well take some time...experienced ICE technicians can still have problems identifying the cause of a simple dash warning light ( “Engine Fault”)....even with all the modern computer diagnostics !
Right - and that exact same skill set (diagnostics from a computer code) is the skill set needed to deal with EV's.
 
This 3rd party Tesla repair service shop seems to have gained a following on YouTube for knowing the Tesla EV's inside out.
This video is absolutely chockablock full of interesting stuff about Tesla EVs and its problems.

I have time-linked it to the part I found the most interesting, this Tesla repair expert says along the lines "Tesla really doesn't want their damaged/past-their-warranty cars on the roads, but if Tesla wants the cars insurance costs to be generally affordable they have to support 3rd party repair/service shops".
https://youtu.be/o-7b1waoj9Q?t=826

Musk’s recent statements of how good his automobiles are for the environment, and how the company was founded to help mitigate climate change, the idea of a single chip rendering a relatively new car nigh unusable doesn’t exactly mesh with one another.
https://www.thedrive.com/news/27945/a-single-component-can-brick-older-teslas-and-tesla-wont-fix-it?xid=twittershare

----------------------------------------------------------------------------------------------------------------------------------------
Q: Does Tesla AP handle tight turns well?
A: No. In this video you can see the vehicle crossover and drive on the wrong side of the road at just 18 mph.
[youtube]HPq_V25cLGY[/youtube]

------------------------------------------------------
Been news about Googles Waymo saying Elon is wrong about LIDAR. Most of the articles are paywalled.
All up I think it's good that we see all these "self-driving tech wars" between these companies, should angry up the blood and finally get affordable self-driving done.
I have watched A LOT of Googles Waymo self-drive demonstrations on YouTube and Waymo actually looks far more smarter in knowing what's going on. In fact, I am sure if Tesla had Waymos self-driving ability, Tesla would claim FSD is now complete/perfect.

I think there is a large difference in general attitude between how Google's Waymo and Tesla believe what Full-Self-Drive should be, that is, Waymo's view of the self-driving car should be truly perfect, in that someone can jump in front of a speeding Waymo car just 3 meters in front and the car will stop in time %100 of the time.
Conversely, I think Elon/Tesla's main goal is, "if the car can just basically compare to a generally bad driver in road-statistics then the FSD is ready."
https://www.adelaidenow.com.au/motoring/googles-waymo-slams-elon-musk-over-claims-its-selfdriving-technology-is-doomed/news-story/5beaf28fc44e3ad618044f3f56f84eea
https://mashable.com/article/waymo-lidar-sensors-elon-musk-tesla/
https://www.businessinsider.com/waymos-execs-defend-lidar-2019-5/?r=AU&IR=T
https://www.heraldsun.com.au/motoring/googles-waymo-slams-elon-musk-over-claims-its-selfdriving-technology-is-doomed/news-story/5beaf28fc44e3ad618044f3f56f84eea

In other words, the goals/standards are very different. I think the overly pure geek 100% perfect attitude of Waymo and also the cost of LIDAR is holding Waymo back from deploying

But the LIDAR costs is well on its way to being solved, manufacturing giant Nikon has announced its plans to "mass produce" a LIDAR sensor. I really think this is big news, this has really held LIDAR back, the costs.
Nikon will help build Velodyne's lidar sensors for future self-driving cars
Mass production is expected to begin in the second half of this year

https://www.cnet.com/roadshow/news/nikon-velodyne-lidar-sensor-manufacturing-self-driving-cars/
Velodyne Lidar partners with Nikon to mass-produce cheaper lidar sensors
https://www.therobotreport.com/velodyne-lidar-partners-with-nikon-to-mass-produce-cheaper-lidar-sensors/
Beyond autonomous vehicles and advanced driver-assistance systems, Velodyne plans to use Nikon’s mass manufacturing scale as lidar sales expand within other emerging markets. The company said the partnership will help its low-cost lidar solutions benefit a range of business segments, including robotics, security, mapping, agriculture, and unmanned aerial vehicles.
----------------------------------------------------------------------------------------------------------------
10,000 Electric Hatchback Pre-Orders "Overload" Volkswagen's IT Systems
https://www.zerohedge.com/news/2019-05-10/10000-electric-hatchback-pre-orders-overload-volkswagens-it-systems
----------------------------------------------------------------------------------------------------------------------------------

I am starting to feel evil jerk posting these charts, but logically, there is nothing I can say/post/do that could actually hurt Tesla.
D6jCLebXoAAmaEV.jpg
 
Hillhater said:
billvon said:
.... While I had a Leaf, the Nissan dealership was quite capable of taking care of maintenance. ......
Maybe , one day , perhaps you will take the time to have a look at the world outside California ! :es:

They maintain them here in most of BC as well.

I meen most of the stuff is tires and windshield wiper blades but they do send techs for training at most of the Nissan shops here.
 
Great video breaking down the cost of a Tesla 3, WSJ claim 40% of the cost of the car is the battery pack. While the story is paywalled the video is free to watch.
The way I see it is, cost means energy, so if the battery cost 40% of the car then that is also where a lot of energy and thus co2 was emitted making the car. For the sake of equivalency, $10,000 worth of petroleum fuel is a lot of miles driving.

Brother, Tesla Can’t Spare a Dime
https://www.wsj.com/articles/brother-tesla-cant-spare-a-dime-11557739800
https://www.wsj.com/video/the-secret-to-why-a-tesla-costs-so-much-hint-batteries/65F3A21D-0837-4DA6-B739-612124815603.html

New article from BI, maybe interesting, a bit click-baity?
It's time for Tesla to go into stealth mode for the rest of 2019
https://www.businessinsider.com.au/why-tesla-should-go-into-stealth-mode-for-the-rest-of-2019-2019-5?r=US&IR=T
--------------------------------------------------------------------------------------------------------------------------------
New article. Tesla's sales numbers for April in China came out, they were incredibly low, so low they didn't make it on the top 10 EV car sales list in China.
May 15, 2019
Slow Sales In China Highlight Tesla's Demand Problem

https://www.forbes.com/sites/jimcollins/2019/05/15/slow-sales-in-china-highlight-teslas-demand-problem/#3482630d46ed
The Model 3 does not show up on Gasgoo’s list of the top-ten selling NEVs in China in April. The number ten model on that list, JAC’s iEV6E, sold 3,002 units for the month of April. So, the only conclusion I can draw (I have reached out to Tesla management for comment; they have yet to respond) is that Tesla did not even sell 3,000 Model 3s in China in April.

That elicited a “whoa” from me. There has been so much hype in so many places about the Model 3 that I had assumed it would come roaring out of the gate in China as soon as models were available there.

--------------------------------------------------------------------------------------------------------------------------------
New Tesla Model 3 Autopilot crash decapitates the driver.
There have been a lot of "suspicious" crashes with Tesla where the car was totally destroyed so the investigators can't work out of autopilot was engaged or not. But this time the car made it through the crash in good condition so the investigators can pinpoint that the autopilot was entirely at fault.

In this case, the driver activated Autopilot, took his hands off the wheel, and was dead 10 seconds later after he hit a tractor-trailer crossing the highway
https://www.ntsb.gov/investigations/AccidentReports/Reports/HWY19FH008-preliminary.pdf
https://www.zerohedge.com/news/2019-05-16/tesla-driver-was-using-autopilot-fatal-march-crash-ntsb
https://www.theverge.com/2019/5/16/18627766/tesla-autopilot-fatal-crash-delray-florida-ntsb-model-3
D6sdoYNXsAAILUm.png

--------------------------------------------------------------------------------------------------------------------------------
T. Rowe Price Associates Inc. sold roughly 81% of its shares of Tesla Inc. over the first three months of the year, marking a sharp retreat for a firm that for years had been one of the electric-car maker’s biggest investors.
https://www.wsj.com/articles/tesla-investor-t-rowe-sells-most-of-its-shares-in-car-maker-11557964482
 
Yup the gigafactory 3 will never happen lol. https://www.teslarati.com/tesla-gigafactory-3-jobs-first-interior-images
 

Attachments

  • Gig 3 800.jpg
    Gig 3 800.jpg
    67.3 KB · Views: 1,922
[youtube]sMvdI8nSAJE[/youtube]

Look at the european autopilot update, not a very calming experience, id rather just drive if someone tells me this car can drive better than me then i laugh at you watch the video and be amazed at a computer child behind the wheel responsible for life crazy this is even allowed if i see a tesla i hang back or go round if safe to.
 
Ianhill said:
Look at the european autopilot update, not a very calming experience, id rather just drive if someone tells me this car can drive better than me then i laugh at you watch the video and be amazed at a computer child behind the wheel responsible for life crazy this is even allowed if i see a tesla i hang back or go round if safe to.

#|1 Keep your hands on the wheel. Its the law.
#2 this will get better on an exponential curve. Every-time a human has to intervene it logs it and send the data to HQ
They have how many cars out there driving gathering free data for them?
I give this 1 year till its smooth for 99.9 % of all driving.
2 years till its 5x a human.
Tesla has for lack of better words found a loophole to get owners to pay them to be beta testers and data collectors!
 
Arlo1 said:
Ianhill said:
Look at the european autopilot update, not a very calming experience, id rather just drive if someone tells me this car can drive better than me then i laugh at you watch the video and be amazed at a computer child behind the wheel responsible for life crazy this is even allowed if i see a tesla i hang back or go round if safe to.

#|1 Keep your hands on the wheel. Its the law.
#2 this will get better on an exponential curve. Every-time a human has to intervene it logs it and send the data to HQ
They have how many cars out there driving gathering free data for them?
I give this 1 year till its smooth for 99.9 % of all driving.
2 years till its 5x a human.
Tesla has for lack of better words found a loophole to get owners to pay them to be beta testers and data collectors!

Problem is europe have made regs that limit the computers steering input, its an update that has caused the issue that takes it a step backwards in useability unfortunately.

I get there will be deaths to make the roads safer overall in the long run it's the only way without a simulated world to test it in that would need to be a perfect replica for the errors learnt to transpose into the real world so only the real world will do really just don't want to see one break the motorway vurge in a panic and kill many innocent that choose not to be part of an experiment that really is a necessity to make a safer future.

People flash the tesla to giveway but it doesn't see it, it's like a user with eyes that sees well but cannot sence all of the danger yet or is it thinking far enough down the road to act in advance thats going to be hard and take alot of time to emulate, I'm not sold on the lack of lidar or the way the ai learns new objects it seems very dull and more distracting than life saving at the moment it's a luxury for the lazy rather than a nessacity that must be implemented to save lives.

When I drive a road I've not been on before I don't drive like a drunk or do I take many times to learn it so I think 5× better in 2 years is ambitious.
Have you watched the ai in a race car that could not beat the human (guy martin) then forgot to warm it's tyres up on out lap and spun went full throttle at a bend lucky to land in gravel trap not really the correct way to beat the human I think. There's so much we feel to race a car at it's edge that it will not beat us for a long time I'd say 10 years till the programming and sensor depth is there for it to be thinking like a race driver and act on instinct.
 
Arlo1 said:
Tesla has for lack of better words found a loophole to get owners to pay them to be beta testers and data collectors!

That's called crowdsourcing. Was done on a massive scale in the software industry more than once. Google did that for their Translate service, and then supposedly some company did an OCR effort via Captchas. I guess Google continues to do that with Maps.
 
Ianhill said:
Problem is europe have made regs that limit the computers steering input, its an update that has caused the issue that takes it a step backwards in useability unfortunately.

So this is a EU issue, glad you recognize this.

People flash the tesla to giveway but it doesn't see it,

You mean people speeding ?

I'm not sold on the lack of lidar or the way the ai learns new objects it seems very dull and more distracting than life saving at the moment it's a luxury for the lazy rather than a nessacity that must be implemented to save lives.

Well, you're the expert. But worth noting that Tesla actually targets safety first, not luxury or laziness. Their approach is incremental due to legal reasons.
 
It will be a fair while before a self-driving car can out-compete an expert human driving operating at the top of their game, but that's not the point. The point of the AI is to be, on average, better than an average human driver on any given day. Considering how many accidents occur due to people simply failing to "drive" by not paying attention, it's a pretty low bar.

There will always be a small number of fatal accidents attributable to AI and this will be a tragedy for the loved ones of those involved, but it will be a much smaller number than would occur by human drivers. This is a hard concept for many people to accept, just like some can't accept that in terms of preventing accidents, human lives have a monetary cost. That there is such a thing as an "acceptable death rate". We grudgingly accept if Uncle Tom killed himself because he fell asleep at the wheel. He shouldn't have driven, he should have stopped when he felt tired, but it was his decision and ultimately he died by his own hand. By comparison, his death due to a computer glitch seems arbitrary and unnecessary. Poor Tom did nothing wrong - it was the incompetence of a programmer somewhere that took him.

cricketo said:
People flash the tesla to giveway but it doesn't see it,

You mean people speeding ?

I think this is unique to the UK, but flashing your headlights at someone can mean different things depending on the situation and context:

[*]For no obvious reason when approaching you means unexpected hazard ahead (horse rider, broken down vehicle, police speed trap)
[*]Signal anger when you just overtook another car in a manner they perceive as being too close to them (i.e. you woke them up)
[*]When behind you on a multi-lane road is a firm request to exercise some f**king lane discipline and let them pass (stop hogging the fast lane)
[*]...and the unusual one: to signal to a waiting vehicle to pull out of a junction or turn across in front of you

So the same flash of the lights can mean "get the hell out of the way - I'm coming through" or "you go first". The "giving way" gesture often occurs at relatively high speed when a vehicle is trying to merge on to a busy road. It's a signal to the merging driver that there isn't a safe gap for them to pull into, but as a courteous fellow motorist you will create one for them at the last moment and so relies on unspoken (and assumed) trust. In other situations, such as congested junctions people use it to reverse the normal right of way (priority) out of a sense of fairness or to try and help keep the traffic moving. This sometimes causes collisions when well-meaning people signal another car to proceed only to be struck by another car who couldn't see the signal.

So, yeah, hardly surprising a first-generation AI designed in California has trouble discriminating this esoteric custom.
 
All fair points raised and good discussion no shit flinging so far, I agree it's a must to stop idiots doing triple figures in 30 zones and progression is needed the death count is unacceptable with humans leading the way.

Flashing lights is not speeding but means be viral caution ahead, or at a junction it can mean I give way so the other car knows you have let them out not speeding and flashing to say out my way on the 3rd lane of motorway lol.

Let's hope that in 2 years it can beat me in all ways I'll happily eat my words many thousands of lives will be saved so i wont choke I'll be in tears we all have lost someone close in a metal box more than likely.
 
Punx0r said:
I think this is unique to the UK, but flashing your headlights at someone can mean different things depending on the situation and context

I've driven quite a bit in Europe, plus some of these signals are practiced in the US as well. I don't think edge case road behavior outside of the traffic rules should be focused on when developing FSD capabilities. There has to be a baseline somewhere - if Teslas were to behave aggressively on the road, that causes for other drivers to do the same which makes it progressively more hard for AI to keep up.
 
Ianhill said:
Let's hope that in 2 years it can beat me in all ways I'll happily eat my words many thousands of lives will be saved so i wont choke I'll be in tears we all have lost someone close in a metal box more than likely.

Elon promised feature complete by end of the year. I know he often makes a promise and delivers 3 years later, but that's typically when it comes to outside vendors, supply chains, and so on. This is purely software, and 100% in-house Tesla. I have no reason to doubt him on this. I don't own a Tesla at this time, but I'm actively looking for a used X so I can get their FSD stuff when it becomes available.
 
cricketo said:
Punx0r said:
I think this is unique to the UK, but flashing your headlights at someone can mean different things depending on the situation and context

I've driven quite a bit in Europe, plus some of these signals are practiced in the US as well. I don't think edge case road behavior outside of the traffic rules should be focused on when developing FSD capabilities. There has to be a baseline somewhere - if Teslas were to behave aggressively on the road, that causes for other drivers to do the same which makes it progressively more hard for AI to keep up.
Yes there has to be a baseline, and infact flashing headlight signals can be an offense in some countries .
But a safe AI system must be able to recognise and respond as necessary to all relavent inputs.
I dont read it as the Tesla “behaving aggressively”... but more about it beian able to interpret a common road signal (information input) and take apropriate action...in the way an experienced driver would.
In order to do that you have to be familiar with local conventions
 
That would require true artifical intelligence in that it would have to be able to infer what others should know i.e. a theory of mind
 
Punx0r said:
That would require true artifical intelligence in that it would have to be able to infer what others should know i.e. a theory of mind

I done an over the air update myself, a new firmware version was written and uploaded in seconds to my non artificial intelligence so i can act in accordance now, i won't flash the tesla driver without eye contact and acknowlegment, ill allow room and assume im unseen, be more vigilant around the ai car, assume its in autopilot and give it room to learn by error over time.

I say treat it like any other learner driver that uses the road, it should be visable to see when in autopilot so im aware It could act up at anytime and give room, but most bully learner drivers thinking they have advanced skills when a lap around the green hell will see them in a barrier.
 
I am glad you guys have a nice long thread hammering on this subject.

After working in the "EV Fast Charge Industry" for a couple years, here are my thoughts.


Model 3
1) My #1 anxiety is around parts, repair, service, firmware, scrapping, etc.

2) My #2 anxiety is around Charging. This is a Trillion Dollar jackpot and big companies are getting bitchy and greedy. Of course, since we are capitalists here in the US, everybody wants to block others and dominate this sector. ... Sell people Juice ... Sell peoples data... its a big deal. ... to cut to the end... NOBODY can handle the job yet. Even if Tesla hired 100 guys to work on it today... there is just too much involved to meet the needs in time. The only hope is for the business-nerds at big companies like Tesla to accept help and open up standards/share load. ... ... I could install a thousand DC fast chargers that work for Model 3 and it would not even be a DROP IN THE BUCKET. We are talking mega-scale here. Giga-scale. Terra-Scale.

There are 168,000 gas stations in the US (really??? I figured there were far more than that!)

There are a quarter billion cars registered in the US
When all those go electric... some number will be able to charge at home.

MANY will not.
For them we have DC Fast Charge, ... because... I highly doubt that the kind of work places which wont even give their employees off-street parking, are going to provide EV Parking...

So...
We are looking at DC Fast Charge...

At something that looks half way between a Starbucks and a Gas Station. Maybe like an old A&W where you would pull in with your car and a girl on roller skates would fill you up..

Anyway... I have no interest in any of that other than keeping in the focal point that my people (poor people) will drive the revolution and we better be ready for them.

A bunch of big hitters at my previous company invested in Model 3s. They bought them. They are currently slow-charging them at work on J1772. Total bullshit.

Yea - the adapters are starting to come out... but if you have ever handled a CHAdeMO you realize that it is perversely insane to run an adapter off of that down to Tesla input. It is also... limiting... as a lot of those CHAdeMO rigs are only good up to 50KW.

CCS1 (and CCS2 over seas) is taking over (even tho it is big and ugly) and making this adapt to a Tesla Connector is also about as lame as a FWD Ebike on a steep dirt hill.

Tesla can play nice... and help/let/encourage other charge companies to run one of their cables. It is really all about the data - so the charge statistics and ongoing reliability programs... and that is like... 10 minutes for a programmer to line up between parties... so no excuses

Alternately - Tesla has to eat it and install a CCS1 inlet as a retrofit into the cars.

... It was an epic move to create the Tesla Sexy-Charger. It was a dick-move to cock-block everyone who is trying to get us to the finish line. I can say that you would see Tesla cables hanging off of "other chargers" if the greed stopped and the honest evaluation of charge-need started.

-methods
 
Punx0r said:
That would require true artifical intelligence in that it would have to be able to infer what others should know i.e. a theory of mind


I vote for super-angry-hive-mind driving.

If a bunch of bugs or birds (with a mind the size of a pin head) can zoom around in a pack/hive/flock. . . then we better be able to.

-methods
 
methods said:
There are 168,000 gas stations in the US (really??? I figured there were far more than that!)
And 48,000 chargers (level 2 and 3.) We have a ways to go - but that's a pretty good start.
There are a quarter billion cars registered in the US
When all those go electric... some number will be able to charge at home.
MANY will not.
Most people will be able to charge at home. Here, even apartment buildings are putting in level 1 chargers (fancy name for an outlet) in their dedicated parking.

It can be a mistake to think that you need a level 2 or level 3 charger to use an EV. The average American spends 48 minutes driving a day. That's 23 hours 12 minutes that the car is parked somewhere. During that time, an ordinary 5-15 outlet can provide 120 miles of range.

Here at my company there was a big hubbub over the small number of level 2 chargers we had. Solution - put in a lot of level 1 chargers (outlets.) That gives people about 40 miles of range during the 8 hours they are parked. It is very rare to need more than that for a trip home.
We are looking at DC Fast Charge...
Level 3 chargers will be important for people taking long trips and applications like long haul trucking. But most people simply won't need them.
A bunch of big hitters at my previous company invested in Model 3s. They bought them. They are currently slow-charging them at work on J1772. Total bullshit.
Why? Sounds like a good solution.
It was an epic move to create the Tesla Sexy-Charger. It was a dick-move to cock-block everyone who is trying to get us to the finish line. I can say that you would see Tesla cables hanging off of "other chargers" if the greed stopped and the honest evaluation of charge-need started.
There are Tesla adapter cables hanging off Chademo chargers (I have one of those adapters.) There are J1772 adapters hanging off Tesla level 2 chargers (I have one of those, too.) Seems to all work out.
 
billvon said:
There are a quarter billion cars registered in the US
When all those go electric... some number will be able to charge at home.
MANY will not.
Most people will be able to charge at home. Here, even apartment buildings are putting in level 1 chargers (fancy name for an outlet) in their dedicated parking.

It can be a mistake to think that you need a level 2 or level 3 charger to use an EV. The average American spends 48 minutes driving a day. That's 23 hours 12 minutes that the car is parked somewhere. During that time, an ordinary 5-15 outlet can provide 120 miles of range.

Its not recommended to rely on a L1 (15a 120Vac outlet) You will wear out the connections and your cable as well its super slow for charging when you really don't want it to be.

I had a mode s for 3 months last winter and I put a drier plug in the garage after 1 week.
The 120Vac charging was at 4km/h So I would only get 32km in the time I needed it to fully charge. I ended up at the super charger in town just to get it up to an acceptable level. Before getting the drier plug installed.
 
Recent survey ( electrical) to estimate the requirements for future EV home charging, in a modern suburb of Melbourne,...concluded that most apartment blocks did not have the necessary capacity to support more than 4 or 5 10Kw charge points ( blocks with 20+ floors, 100+ residents) . The buildings electrical supply and distribution hardware would need major upgrades (replacement), but also the neighbourhood substations and power feed was considered unsuitable for significant % of EV charging loads.
Obviously even many of the newly developed sub divisions are built to a “Minimum “ spec for cost reasons, and face having to consider complete reworking of the network power facilities to cater for future EV uptake.
 
Arlo1 said:
Its not recommended to rely on a L1 (15a 120Vac outlet) You will wear out the connections and your cable as well its super slow for charging when you really don't want it to be.
We have 42 Level 1 chargers, in use 8-10 hours a day, 250 days a year. No "wearing out" the connections. You have to use decent outlets, of course, not Home Depot $0.49 specials. And slower charging is better for batteries anyway.

Again, the only thing that matters is that you get enough charge to get where you are going. And 40 miles a day added at work is enough for everyone here to get home. Want to charge faster? Go for it. But it's a "nice to have" not a "need" for people who charge at work / at home for their commute. (Of course we still have people who get mad when they can't get one of the L2 chargers. Human nature.)
 
Hillhater said:
Recent survey ( electrical) to estimate the requirements for future EV home charging, in a modern suburb of Melbourne,...concluded that most apartment blocks did not have the necessary capacity to support more than 4 or 5 10Kw charge points ( blocks with 20+ floors, 100+ residents) .
Much better to put in 35 level 1 chargers (same power, much cheaper.) That supports up to a third of the residents having EV's - and they get 80 miles worth of charging a day.

Need more? Put in a smart grid that supports all the parking spots with level 2 charging, and is boosted by solar. During the day you charge at full rate, and at night it backs off to a level that the existing infrastructure can support.
 
Status
Not open for further replies.
Back
Top