Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable

[EVMeister]

Can you send a link where I can read about such radar? do you know the price?

Yes. I give you link. $5000.

 

[StarFoxisDown!]

BTW., 85,337 VINs registered in Q1 so far means the following:

• Even if Tesla stops filing for new VINs today, with the "simple 85% rule" this would suggest a Q1 production target of 72,500 Model 3's.
• But there's still 48 days left from Q1, and in Q4 Tesla registered another ~9,000 VINs in the remaining 48 days of the quarter. If Q1 ist similar, that means ~94,300 VINs in Q1, which gives a Q1 production target of 80,150 Model 3's.
• But if we use my full estimation method that worked well in Q4, then it's going to be 97,900 VINs, which gives a Q1 production target of 83,200 Model 3's.

So after Q4 Model 3 production of 61,400 this is a pretty bullish increase for Q1 in the +18-37% range, and most of those units are expected to be high ASP configurations like in Q3, but using Q1 production efficiencies (!).

The usual caveats apply: Tesla might depart from the pattern. Tesla didn't give any Q1 production guidance.

I'll keep updating my estimate as the quarter progresses.

Tesla guided to around 10k Model 3's being in transit for Q1 right? I'm still trying to get my head wrapped around the tiny profit for Q1. If there were 10k in transit, with the higher level of production and with the much higher ASP mixture, shouldn't the profit still be close to Q3 and higher than Q4? The only wild card is S/X production levels. Even if production and deliveries of S/X is like 15-17k, with the higher revenue and profit on each one, that should make up for some of the loss of lower S/X production. Anyone thinking Elon was being wildly dramatic in the guidance of a small profit?

 

[Heimdallr]

BTW., 85,337 VINs registered in Q1 so far means the following:

• Even if Tesla stops filing for new VINs today, with the "simple 85% rule" this would suggest a Q1 production target of 72,500 Model 3's.

Too lazy to look it up, does the "simple 85% rule" as you call it work for all previous quarters? I am really encouraged by the high number of VINs registered.

 

[Fact Checking]

An easy way to do this would be with dual-antenna interferometry (one antenna on the left side of the car, one on the right), to create a large virtual aperture without needing a physically large antenna.

Radar interferometry was my guess too when the "second radar" speculation started based on the two radar channels diagram from Tesla.

Note that a left-right in-bumper placement would mainly improve horizontal angular resolution, while potholes need vertical resolution.

Two radars would also increase hardware redundancy: radar is one of the most important crash prevention sensors, and it's also weather invariant, so improving it is probably a priority for Tesla.

 

[Krugerrand]

Damn, I was sure it was “Hawaiian Roller Coaster”.

Context, Maximus, context.

 

[Krugerrand]

Mine was winninger.

It was a pretty darn good attempt, but in then end it was the wrong song.

 

[Fact Checking]

Too lazy to look it up, does the "simple 85% rule" as you call it work for all previous quarters? I am really encouraged by the high number of VINs registered.

Yes, the "85% of registered VINs" method worked well for Q2, Q3 and Q4 of 2018, and I used it to predict Q4 production of 62,800 units, which was close to the reported 61,394 units, an error of margin of ~2.3%, a slight over-estimate.

(The usual caveat applies: past performance is not indicative of future results.)

 

[KarenRei]

Can you send a link where I can read about such radar? do you know the price?

I don't think there's any multi-antenna radars on the market designed for vehicular use at present. Well... okay, they are used for vehicle-mounted ground penetrating radar tasks, but that's not the same thing And also if you want to get technical, since vehicle radars generally phased array, they're all multi-antenna... it's just that all of the antennas are adjacent to each other, so they don't get you an aperture boost Making an "off the shelf" multi-antenna system would be more challenging than a single antenna, because the distance between the two antennas must be known with high precision. But a manufacturer like Tesla could certainly accomplish it if they wanted high radar resolution.

As for using radar to measure roughness or smoothness on the scale of the wavelength... ever seen a radar map made by a satellite? That's exactly what they're doing E.g.:

That's a radar map of Titan from Cassini. The black is seas of liquid methane, so it's smooth. The various shades of white are the landscape - the rougher (on the scale of the wavelength), the brighter. It's due to how the signal reflects - if the target is smooth, it tends to reflect away from you, while if it's rough, some fraction reflects back at you. But bumpiness on a scale significantly smaller than that of the wavelength doesn't affect the reflection.

Now, satellites use SAR (Synthetic Aperture Radar), because on the downside, they're very small relative to the distances that they have to measure from, but on the upside, they're moving extremely fast (SAR creates a large virtual aperture not by having a large spacing between multiple antennas, but rather relying on the fact that by the time the signal returns, the satellite has moved, so a single transmitting antenna can act like second receiving antenna at a significant distance away. For vehicular uses, you'd use two physically separated phased arrays. But the physics is the same.

 

[mongo]

Radar interferometry was my guess too when the "second radar" speculation started based on the two radar channels diagram from Tesla.

Note that a left-right in-bumper placement would mainly improve horizontal angular resolution, while potholes need vertical resolution.

Two radars would also increase hardware redundancy: radar is one of the most important crash prevention sensors, and it's also weather invariant, so improving it is probably a priority for Tesla.

Although... Elon has been talking video instead of radar (guess he isn't as worries about spaceships landing in sandstorms).

Applying a hydrophobic coating to the radar (located just below nose of car) should help. Easy to do yourself or Tesla service can do it. We’re also working on vision-only driving.

Elon Musk on Twitter

@KarenRei is right on the parallax, my city driving in a truck does not scale to HW speeds in a sedan. The NN would need to pick up the pothole from still frames,

 

[Mark's Beagle]

BTW., 85,337 VINs registered in Q1 so far means the following:

• Even if Tesla stops filing for new VINs today, with the "simple 85% rule" this would suggest a Q1 production target of 72,500 Model 3's.
• But there's still 48 days left from Q1, and in Q4 Tesla registered another ~9,000 VINs in the remaining 48 days of the quarter. If Q1 ist similar, that means ~94,300 VINs in Q1, which gives a Q1 production target of 80,150 Model 3's.
• But if we use my full estimation method that worked well in Q4, then it's going to be 97,900 VINs, which gives a Q1 production target of 83,200 Model 3's.

So after Q4 Model 3 production of 61,400 this is a pretty bullish increase for Q1 in the +18-37% range, and most of those units are expected to be high ASP configurations like in Q3, but using Q1 production efficiencies (!).

The usual caveats apply: Tesla might depart from the pattern. Tesla didn't give any Q1 production guidance.

I'll keep updating my estimate as the quarter progresses.

I am guessing Tesla has moved to a under promise, over deliver model moving forward.

 

[Krugerrand]

Funny, in US at first Tesla service centers cloaked the VINs on the cars. I think it was because VINs were emailed to reservation holders and they did not want too much attention until the 3's were ready for delivery. Or maybe there was some other reason....

Different circumstances. They’re using codes and readers on the ships to track inventory. If they were hiding those, keeping track would be logistics hell.

We could just chalk it up to non-Americans not stalking Tesla or being as rabid.

 

[Fact Checking]

As for using radar to measure roughness or smoothness on the scale of the wavelength... ever seen a radar map made by a satellite? That's exactly what they're doing

Note that I believe that beyond retrieving the strength (amplitude) of radar reflections, high end satellite radar systems are also able to retrieve any shift in the polarization of the reflected photons.

This allows the sensing of additional surface attributes: different surface materials and textures will affect polarization differently. This allows the extraction of (some) surface features whose characteristic size is well below radar wavelength resolution.

For automotive use polarization might allow the disambiguation of metal vs. concrete reflections.

But that's just speculation - in satellite based remote sensing it's mainly used to categorize type and depth of foliage, IIRC. Haven't seen any specific research on this related to automotive use.

 

[StarFoxisDown!]

Also was watching the Power Lunch segment today and just constantly rolling my eyes. Their one-sided bias is so obvious. It's also obvious they read from a script which is whatever the top 5 FUD articles from the week. They can't get basic understanding that Tesla front loaded delivery staff during the US Model 3 launch. A child could understand the logic behind the layoffs.

My personal joy is knowing that these same guys were wrong about Amazon.....year after year after year. Constant bashing of Amazon saying it's multiple/valuation was too high and it won't make money.

 

[M3Rider]

Just as you have a memory of the road a FSD car will have the memory of every other FSD car that shares the same hardware platform. It will know what happened to a car 1 minute in front. The memory will be compared with what the car sees and positioning data before every single manoeuvre.

Main cameras are behind the windscreen so they will be kept clean at all times. Radars see through fog and snow. Cameras see through the dark. If a camera is covered with snow the car can let you know so you can wipe it before setting off. Having direct heating around some cameras might help to.

Snow will be problematic but far from impossible to deal with just as it is for us mere humans. I'm looking forward to "donut mode" in my FSD Tesla.

I do not believe that Tesla will be accumulating the memory of recent road conditions.

I think you can't count on road to be visible or for another Tesla to provide you what it saw etc. ahead of time.

The assumption needs to be that it's been snowing all night and you're sending your driverless FSD car to pick up someone in the morning and the roads have not yet been cleared in places, so your car will be first to navigate them.

If it can't do it, then you can't satisfy your customer's taxi request.

We all speculate what will happen. There were some vids posted. My thought about those videos is that 90% success rate is far from 99.999%.
With AP the liability lies with the driver.
If Tesla assumes FSD liability and it has a 90% success rate, then they'll go bankrupt due to the 10%. It needs to be much much better. Pretty much no mistakes.

 

[KarenRei]

OT

Sorry for the offtopic, but I just wanted to note (since Americans spent so long talking about the Superb Owl in this thread): it's increasingly looking like Iceland is going to troll Eurovision again this year - perhaps not quite to the degree that we did with Silvía Night, but close. The band "Hatari" (Hater) handily won yesterday's semifinals, with the song Hatrið Mun Sigra ("Hatred Will Win"). Their stage performance is in full S&M gear, including a band member in a gimp outfit who gets abused by the lead singer (the band is known for approaching the stage by physically shoving the audience out of the way). While Eurovision songs are generally of the "Tolerate everyone!", "Love is awesome!" and "Can't we all just get along?" variety, the chorus of their song proudly declares "Hatred will win, Europe will collapse" and declares love dead.

IMHO, their music sucks (not my taste), but I'll be cracking up nonstop watching them up on stage in Tel Aviv (assuming that they win the final on 3 March). And besides, most Eurovision music sucks regardless. Oh, and they're already driving our social conservatives into fits, which wins anyone kudos in my book .

 

[SteveG3]

The Canaccord upgrade is exactly what we anticipated as the Tesla story becomes obvious to people paying less attention than us.

This is the start of the Great Awakening. Investors, analysts, and financial reporters rubbing the sands from their eyes on TSLA.

There will always be TSLA bears and shorts, but we are about to enter a period in which there are more TSLA bulls and longs. That is a new world for us.

I hope your view proves to be more accurate on this than mine, but, for years I’ve been saying I find it most probable,

- the vast majority of the flood of ‘foolish’ talk re Tesla in the media and among analysts (and even major “bears” such as Chanos) is not about these people believing ‘foolish’ things, but rather that their current employment requires they’re trying to persuade the public to believe ‘foolish’ things

- these efforts are unlikely to largely dissipate until about 2022 at the earliest. until then, I see TSLA continuing to trade at 10% to 50% below fair value.


Again, I’d genuinely be quite pleased if it turns out you’ve got this more accurately than I do and analysts and financial reporters experience a great awakening. Perhaps reality will turn out to be somewhere between our views.

 

[KarenRei]

Note that I believe that beyond retrieving the strength (amplitude) of radar reflections, high end satellite radar systems are also able to retrieve any shift in the polarization of the reflected photons.

This allows the sensing of additional surface attributes: different surface materials and textures will affect polarization differently. This allows the extraction of (some) surface features whose characteristic size is well below radar wavelength resolution.

For automotive use polarization might allow the disambiguation of metal vs. concrete reflections.

But that's just speculation - in satellite based remote sensing it's mainly used to categorize type and depth of foliage, IIRC. Haven't seen any specific research on this related to automotive use.

Indeed - AFAIK, this would be brand new ground, if Tesla were to do it. But it would be extremely useful new ground, and is already in use in non-automotive fields (aka, no "revolutionary breakthroughs" required). A lot easier than a lot of the things people are trying to do make better LIDARs, IMHO.

 

[Oil4AsphaultOnly]

Tesla guided to around 10k Model 3's being in transit for Q1 right? I'm still trying to get my head wrapped around the tiny profit for Q1. If there were 10k in transit, with the higher level of production and with the much higher ASP mixture, shouldn't the profit still be close to Q3 and higher than Q4? The only wild card is S/X production levels. Even if production and deliveries of S/X is like 15-17k, with the higher revenue and profit on each one, that should make up for some of the loss of lower S/X production. Anyone thinking Elon was being wildly dramatic in the guidance of a small profit?

They'll be missing (10,000 cars * $55k/car) $550 million in revenue, while still needing to account for the COGS for those cars in-transit. I think the lack of profitability is pretty clear.

What's not stated is that we can probably expect something close (lower ASP due to changing sales mix - maybe $500 million?) as profit for q2.

 

[uselesslogin]

Looks like official short interest is down to 24.7 million. I'm really surprised how much covering has happened while keeping the price down.

 

[Dan Detweiler]

That wasn’t the song, but one bonus point to you for putting two socks of the same color on today.

How do you know?

Dan