Actually its 250 (including DeepScale) AP engineers. Secondly all 1,500 Waymo employees are not engineers. Its just 1,500 who work at Waymo
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Profound progress towards FSD
- Thread starter Grifon
- Start date
Actually its 250 (including DeepScale) AP engineers. Secondly all 1,500 Waymo employees are not engineers. Its just 1,500 who work at Waymo
Lets step back for a second. This isn't even a rewrite.. They didn't have prediction networks to BEGIN WITH...
You can't call something a rewrite of something you didn't even have to begin with.
Secondly previous Tesla have used old public NN architectures of Google. For example using the public inception GoogleNET architecture from 2015. Then they recently moved to ResNET, etc.
I could go on and on about how Google's Deepmind has led every NN innovation and revolution.
This is 100% incorrect.
There are 4 categories in autonomous driving development.
They are perception, mapping, prediction and planning.
If i told you to evaluate a system to see if it meets the criteria for each category..
You wouldn't look at a system with just vehicle (current lane) cut-in detector and say it meets the criteria of knowing all the probable intent, actions, current & future trajectory of every actor in relation to anywhere on road.
Its not oxymoron because this isn't word play.
Last edited:
powertoold
Active Member
Maybe it doesn't fit into the lidar camp's definitions, but the traffic control feature currently does all of the above.
Making money is a poor metric of superiority when it’s largely a product of fooling people it’s a worthwhile option on their cars.
boonedocks
MS LR Blk/Blk 19”
Via FB Tesla Tips & Tricks / Denis Y.
https://i.imgur.com/lXPPbAE.jpg
https://i.imgur.com/SK3DeMV.jpg
https://i.imgur.com/noUxuhy.jpg
willow_hiller
Well-Known Member
He said a year or so for roundabouts worldwide, as they're primarily a concern in Europe.
Unless by "cross country" you literally meant across multiple countries.
Selling emissions credits is what makes money for Tesla, not autopilot. Remove those credits and the money starts draining again.
bd7349
Member
I take that to mean crazy complex roundabouts like you see in Europe. He said it’ll be able to handle roundabouts just not all of them perfectly. The simple ones we have here in the states should be pretty easy compared to what it’ll have to handle in Europe and other countries.
goRt
Active Member
Progressing towards FSD allows Tesla to release cash from the monies already paid for FSD.
Buckminster
Well-Known Member
Umm Waymo need help presumably - design the road around the car... Scalability is off the charts...
An Alphabet company is designing a road for autonomous cars in Michigan
So this is why you don't think FSD is a solved problem @diplomat33 ?
An Alphabet company is designing a road for autonomous cars in Michigan
So this is why you don't think FSD is a solved problem @diplomat33 ?
diplomat33
Average guy who loves autonomous vehicles
If you are implying that Waymo is helping with this project because they cannot solve FSD, no I don't think so. This project is separate from Waymo's own efforts to solve FSD. This is a project by the State of Michigan. There are those who believe that our poor and inconsistent road infrastructure is making autonomous driving harder than it needs to be. So they argue that perhaps we should try to design roads that are more friendly to autonomous cars. That is what this project is all about. Waymo is helping with the project because they are a leader in FSD so they certainly have expertise to contribute in terms of how to make roads more friendly to autonomous cars. There is no doubt that if we had special roads designed just for autonomous driving, that we could make autonomous driving a whole lot easier, that's for sure.
But in terms of why FSD is not "solved" yet, the perception part of autonomous is solved with camera+radar+lidar+HD maps. And Waymo has solved the perception part. Waymo also has advanced planning/driving policy, hence why they are able to deploy robotaxis with no drivers in geofenced areas. The part that is not solved yet is the prediction part of autonomous driving. Predicting the behavior of other vehicles, pedestrians, especially when there are a lot of them, can be a challenge. Consider that some cars forget to signal, some pedestrians change their minds at the last moment, cyclists may forget to signal etc... So you need to predict possible behavior of agents that don't always signal their intent and sometimes behave in an unpredictable way. So "solving" the prediction part may not be possible. But the goal is to train the autonomous car to at least anticipate the possible behaviors of all agents as accurately as possible so that the car can drive as safely as possible.
I base this on what Waymo's CTO said. Last year, in an interview with TechCrunch, he said that:
"Detecting different objects, understanding what they are, classifying pedestrians, cyclists, cars etc... It gets super interesting beyond that. That's step 1. Self-driving 101, you have to see the world and understand the different types of objects. Really where it gets interesting and where we have been doing a lot of fundamental work on the research front and the engineering, is reasoning about the world at the scene level, not just the per object level, and really deeply understanding the interactions between the different objects in the world. How does one car fit with the pedestrian or the cyclist? How do you fit in to that? This very deep, very dynamic, very social interaction, is where a lot of really interesting work happens."
Source:
Lastly, I would add that "solving" L5 is exponentially harder than "solving" L4. That's because L5 involves an absurd amount of edge cases and driving cases to deal with. Getting the autonomous car to handle ALL the cases in L5 reliably is a monumental challenge. "Solving" L5 may not be possible, at least not right away. Waymo has FSD that works but if they deployed it everywhere, it probably would not be reliable enough to remove the driver because of how many different cases it would need to be able handle. That's why Waymo focuses on L4. With L4, you can limit the cases you need to deal with. Less cases makes it easier for the FSD and it makes it easier to validate when your FSD is ready to be deployed without a safety driver. Hence, why Waymo has been able to remove the safety driver for some geofenced routes.
Last edited:
Although I do agree that the emission credits was what made the difference in Q2 earnings, they did sell many vehicles. Many people bought a Tesla for a variety of reason: performance/driving experience, supporting "the mission", and autopilot/FSD. Having autopilot as standard in all vehicles (except for the base SR model 3) is part of the reasons people would buy a Tesla (well, at least the ones that don't know about Comma AI Open Pilot). My take on it is that the folks that work on autopilot did help Tesla sell some vehicles and thus help Tesla make money.
Ummm no. They get the cash immediately. They "recognize" the income on their accounting statement as they progress to FSD. Big difference.
diplomat33
Average guy who loves autonomous vehicles
How do people read this tweet? Is Elon saying 10B km total or 10B km starting from when they release FC wide? And is Elon saying 10B km until L5 or until the nags are removed? What?
Honestly, the regulatory approval thing seems like nonsense to me. Waymo does not have anywhere near 10B km of real world driving and yet they are allowed to deploy driverless robotaxis in AZ. The CA DMV does not require 10B km to approve FSD. The CA DMV will let you deploy autonomous cars on their public roads as long as you register the VINs of the vehicles and file disengagement reports every year. There are lots of companies who have deployed autonomous cars in CA and they file disengagements reports every year and none of them have 10B km of real world driving. The fact is that if Tesla had autonomous driving, they could deploy autonomous cars today in CA if they simply followed those requirements. So it is not out of Tesla's hands. Tesla simply refuses to file a disengagement report on any of their autonomous testing (except the one demo because they were forced to report it).
willow_hiller
Well-Known Member
Does the legislation concerning autonomous vehicles allow the sale and operation of vehicles to the public? Or is all of the current legislation phrased to only allow testing by employees?
diplomat33
Average guy who loves autonomous vehicles
I think it depends on the State. 16 States do allow "full deployment" while others only allow testing.
Here is what I could find with a quick Google search:
"There are currently 38 states along with the District of Columbia that have enacted legislation or issues executive orders regarding Autonomous Vehicles (AVs). 12 states authorize testing, while 16 states and the District of Columbia authorize full deployment."
| GHSA
"The Federal Autonomous Vehicle Policy includes no new rules or regulations—only guidance—for states. Although it will likely be many years before fully autonomous vehicles see widespread deployment, there is much for states to accomplish in areas such as traffic enforcement, insurance, registration, licensing and more."
Regulating Autonomous Vehicles.
DanCar
Active Member
Arizona, Florida, Texas are states where the laws do not prohibit driverless. I don't like how Musk implies it will take a while to get regulatory approval, when none is needed in 16 states.
It's a nice thing to hide behind when you are delayed.
Many companies also use COVID-19 as an excuse to be lazy and slow.
goRt
Active Member
What I said is correct, you've restated it.
The correct accounting practise is the accrue the revenue pending functional milestones, you can also accrue the cost.
Elon is clearly driving the "boss build". This build has a UHD-mapped route from his home to his parking space, running 2020.32.X, with a bit of eye candy added to make it seem "alpha"
Similar threads
