Autonomous Systems

Would you ever own an autonomous vehicle?

  • Hell yeah Bring on our robotic overlords!

  • Fuck you! I'll keep my Indepenence


Results are only viewable after voting.

Tuco

I got Tuco'd!
<Gold Donor>
45,395
73,466
Long term I would be surprised if some of these ride hailing services just become free. Watch a 30s advertisement every 10 minutes or something of driving and they can earn the cash back for the ride. It's not really feasible with gas powered vehicles, maintenance costs are simply too high but electric vehicles can be incredibly reliable and require little maintenance because they're so simple to build in comparison and have fewer parts that can fail.
They have rent-a-bike stations that cost $8 for a half hour of use:
BCycle - Home

Maybe when we get to Star Trek levels of tech we'll have free autonomous taxi-service, but don't expect it to be cheap for a long time. Here's the truth: taxi drivers aren't expensive, uber drivers less so. Autonomous taxis are an amazing goal built on technology that doesn't exist yet. Don't expect it to be commoditized for a long time.
 
  • 1Like
Reactions: 1 user

Tripamang

Naxxramas 1.0 Raider
5,186
31,688
You can already get a free cabride to any of the major tittie bars from anywhere in Vegas during certain hours. The principle already exists but those costs would have to be DRIVEN DOWN, they won't come down by themselves.

It'll prolly be closer to getting free rides to a Best Buy during Black Friday or Target/Walmart promoting free rides to their stores to compete with e-retailers.

This is an awesome idea actually, I could see high end retailers having a store vehicle they can send around to pick up clients.

They have rent-a-bike stations that cost $8 for a half hour of use:
BCycle - Home

Maybe when we get to Star Trek levels of tech we'll have free autonomous taxi-service, but don't expect it to be cheap for a long time. Here's the truth: taxi drivers aren't expensive, uber drivers less so. Autonomous taxis are an amazing goal built on technology that doesn't exist yet. Don't expect it to be commoditized for a long time.

Ford is saying 2021 is the year they'll have fully autonomous cars available for sale, am I missing something when you keep saying it'll be far in the future?

Ford Targets Fully Autonomous Vehicle for Ride Sharing in 2021; Invests in New Tech Companies, Doubles Silicon Valley Team | Ford Media Center
 
  • 1Like
Reactions: 1 user

Lleauaric

Sparkletot Monger
4,058
1,823
This is completely false. Lawn care is a pretty complex problem and few people will pay big bucks to take care of a subset of that problem with a robot.

It's like saying the existence of the roomba has removed the market for vacuum cleaners.

I would say lawns are much different from roombas. Especially if you already don't mow your own lawn. And lawn care isnt that complicated, at least mine isn't. At the root of it just making sure the NPK is straight and hitting it with lime 2x a year and weedkiller once in a while.

Vacuuming the house isn't as big a pain as running a vacuum. And the other part is you have to buy a roomba. Its not worth the purchase. But if it was mowing the lawn, you are simply paying less for what you were already getting.
 

Aldarion

Egg Nazi
8,894
24,274
This wont happen in my lifetime, so its an academic debate, but an interesting one.

When it does become a reality, half the country is going to be totally happy to sign over all their independence to the DMV and will push for outlawing manual driving.

And the other half of us arent going to be able to live in the same country with those soulless fucks anymore.
 
  • 1Like
Reactions: 1 user

Tuco

I got Tuco'd!
<Gold Donor>
45,395
73,466
I would say lawns are much different from roombas. Especially if you already don't mow your own lawn. And lawn care isnt that complicated, at least mine isn't. At the root of it just making sure the NPK is straight and hitting it with lime 2x a year and weedkiller once in a while.

Vacuuming the house isn't as big a pain as running a vacuum. And the other part is you have to buy a roomba. Its not worth the purchase. But if it was mowing the lawn, you are simply paying less for what you were already getting.
Nothing you said even addressed what I said.
This is an awesome idea actually, I could see high end retailers having a store vehicle they can send around to pick up clients.



Ford is saying 2021 is the year they'll have fully autonomous cars available for sale, am I missing something when you keep saying it'll be far in the future?

Ford Targets Fully Autonomous Vehicle for Ride Sharing in 2021; Invests in New Tech Companies, Doubles Silicon Valley Team | Ford Media Center
For SAE4, maybe, under certain constraints. 2021 is the kind of date that is ambitious and near enough to be impressive, but not near enough that people will hold you to it.

But SAE5? Not a chance.

http://www.sae.org/misc/pdfs/automated_driving.pdf



The problem with automation is that the difficulty revolves around the amount of constraint in the problem. The more constrained your problem is the more tractable it is and the fewer corner cases you have to worry about. Vehicular autonomy is an exercise in navigating through corner cases, and getting to 100% autonomy is extremely difficult.

Said different, going from 99% to 100% autonomy might be more challenging than going from 1% to 99%.
 

khorum

Murder Apologist
24,338
81,363
They have rent-a-bike stations that cost $8 for a half hour of use:
BCycle - Home

Maybe when we get to Star Trek levels of tech we'll have free autonomous taxi-service, but don't expect it to be cheap for a long time. Here's the truth: taxi drivers aren't expensive, uber drivers less so. Autonomous taxis are an amazing goal built on technology that doesn't exist yet. Don't expect it to be commoditized for a long time.

Nah that's already reduced to an information-theory problem... once you can quantify every factor in an economic relationship, it becomes just another matchmaking application. That's probably why Uber is all over it.

All it really comes down is whether the cost of subsidizing for someone's free ride exceeds the benefit of bringing them to the store and the profit off their purchases. If Target knew that a customer was going to spend 200 bucks on groceries and they were only a $4 uber drive away---they'll HAPPILY commit to providing that ride rather than lose that customer to Amazon Prime.

Then there's the aggregate value-add of customer satisfaction, impulse purchases etc. Uber can probably pitch this to Walmart and Target and walk away with a deal pretty easily.
 

Tuco

I got Tuco'd!
<Gold Donor>
45,395
73,466
Here's one of the fun problems that Ford is dealing with, snow:
The Clever Way Ford’s Self-Driving Cars Navigate in Snow

What they're doing is cool and textbook. They use their velodynes to build up a 3d model of the terrain over time, then do Simultaneous Localization and Mapping (SLAM), which is to match what they['re seeing now to what they saw before, to understand very accurately where they are and theoretically what's around them. I made something very similar for a very different navigation problem last year.

This is basically the only way to do it and is similar to how humans drive. But you know how when there's the first big snow storms and you see accidents fucking everywhere? That's because it's a big fucking problem with no easy solutions.
 

Tuco

I got Tuco'd!
<Gold Donor>
45,395
73,466
Nah that's already reduced to an information-theory problem... once you can quantify every factor in an economic relationship, it becomes just another matchmaking application. That's probably why Uber is all over it.

All it really comes down is whether the cost of subsidizing for someone's free ride exceeds the benefit of bringing them to the store and the profit off their purchases. If Target knew that a customer was going to spend 200 bucks on groceries and they were only a $4 uber drive away---they'll HAPPILY commit to providing that ride rather than lose that customer to Amazon Prime.

Then there's the aggregate value-add of customer satisfaction, impulse purchases etc. Uber can probably pitch this to Walmart and Target and walk away with a deal pretty easily.
Right, and the kind of commoditization that will allow that is farther away than the people selling those ideas want you to think. It's very cool stuff and will happen eventually, just don't get anxious when it's not here in a decade.
 

Tripamang

Naxxramas 1.0 Raider
5,186
31,688
For SAE4, maybe, under certain constraints. 2021 is the kind of date that is ambitious and near enough to be impressive, but not near enough that people will hold you to it.

But SAE5? Not a chance.

http://www.sae.org/misc/pdfs/automated_driving.pdf



The problem with automation is that the difficulty revolves around the amount of constraint in the problem. The more constrained your problem is the more tractable it is and the fewer corner cases you have to worry about. Vehicular autonomy is an exercise in navigating through corner cases, and getting to 100% autonomy is extremely difficult.

Said different, going from 99% to 100% autonomy might be more challenging than going from 1% to 99%.

Ahh you're talking about level 5, okay that makes sense now. Yea I agree that's a long way off, I just think most people would be ecstatic with level 4. If it can get me from A->B within a city and between major cities on it's own I'd be lining up to buy one.
 

Tuco

I got Tuco'd!
<Gold Donor>
45,395
73,466
Ahh you're talking about level 5, okay that makes sense now. Yea I agree that's a long way off, I just think most people would be ecstatic with level 4. If it can get me from A->B within a city and between major cities on it's own I'd be lining up to buy one.
Even level 3 is amazing and will make a huge impact for a lot of people. I got to try bosch's super cruise (level 3) system last year and I can't wait to get one. Being able to just get o nthe freeway, hit a button and resume control when I'm near my destination will be a game changer.
 

Lleauaric

Sparkletot Monger
4,058
1,823
I dont know Tuco, I agree we arent going to see it in the next 5 years... but after that? People re really aggressively pushing to get a handle on this market before anyone else.

Check this out...

Uber's first self-driving cars will start operating within a month

Otto, founded by ex-Googlers, is bringing self-driving technology to trucks

Uber is pushing this really hard. I think they see themselves as the next Apple or Microsoft of transportation. And hey... gaining a Google like control of the nations $700 billion dollar trucking industry seems like a pretty nice market to corner.
 

Tuco

I got Tuco'd!
<Gold Donor>
45,395
73,466
Autonomous trucks are a fun problem. The big challenge with breaking into them is that for passenger vehicles just being able to minimize the participation of the driver is great, but for autonomous trucks the goal is to remove the driver from the vehicle entirely. Adding levels of driver assistance is all well and good, but you're not going to save much in salary by telling the driver, "You can sleep on the job as long as you wake up with this alarm goes off...".
 

khorum

Murder Apologist
24,338
81,363
fuck that. when i can no longer drive then i will take one of these russian roulette machines. for now, me and my pre computer age convertible got some more miles to go.

It's not like they're gonna ban you bro. They'll just have "MEATBAG-ONLY" lanes where the speed limit is a slovenly 80mph and you have to watch people doing pilates or playing videogames at 150mph a couple lanes over.
 
  • 2Like
Reactions: 1 users

Tuco

I got Tuco'd!
<Gold Donor>
45,395
73,466
It's not like they're gonna ban you bro. They'll just have "MEATBAG-ONLY" lanes where the speed limit is a slovenly 80mph and you have to watch people doing pilates or playing videogames at 150mph a couple lanes over.
This is exactly right. I don't think we'll ever drop passenger control over vehicles, it'll just become either discouraged or foolish. Ex: how it was in iRobot.
 

Lleauaric

Sparkletot Monger
4,058
1,823
Autonomous trucks are a fun problem. The big challenge with breaking into them is that for passenger vehicles just being able to minimize the participation of the driver is great, but for autonomous trucks the goal is to remove the driver from the vehicle entirely. Adding levels of driver assistance is all well and good, but you're not going to save much in salary by telling the driver, "You can sleep on the job as long as you wake up with this alarm goes off...".

No doubt.. but whoever solves that problem is going to be richer than fuck.

Another thing. I imagine Fed Ex looks at this like existential threat. If another company comes along and can promise faster delivery for less money they go bye bye.
 
Last edited:

Tripamang

Naxxramas 1.0 Raider
5,186
31,688
Autonomous trucks are a fun problem. The big challenge with breaking into them is that for passenger vehicles just being able to minimize the participation of the driver is great, but for autonomous trucks the goal is to remove the driver from the vehicle entirely. Adding levels of driver assistance is all well and good, but you're not going to save much in salary by telling the driver, "You can sleep on the job as long as you wake up with this alarm goes off...".

I've read that they plan to make the autonomous trucks go city to city on their own, but they would park themselves at depots where regular drivers would take over for the city movement.

I dont know Tuco, I agree we arent going to see it in the next 5 years... but after that? People re really aggressively pushing to get a handle on this market before anyone else.

Check this out...

Uber's first self-driving cars will start operating within a month

Otto, founded by ex-Googlers, is bringing self-driving technology to trucks

Uber is pushing this really hard. I think they see themselves as the next Apple or Microsoft of transportation. And hey... gaining a Google like control of the nations $700 billion dollar trucking industry seems like a pretty nice market to corner.

Whoever ends up owning the robots that do the work, are going to be the only people who end up rich when automation plays out.
 

Tuco

I got Tuco'd!
<Gold Donor>
45,395
73,466
Another cool part about autonomy and robots. The core sensor in all of these is lidar, which are the big spinney things you see on most autonomous vehicles. The big one, the velodyne hdl64e, is $80,000. The smaller ones Ford is using now are the velodyne hdl32s which are 30k. Velodyne released some smaller ones for 8k. Ibeo's range from 12-20k.


There's some very cool developments on that front from quanergy and velodyne to come up with solid state lidar:
Quanergy Announces $250 Solid-State LIDAR for Cars, Robots, and More

The tech behind this is really cool. They use arrays of transistor to direct beams in a very controlled fashion. The capabilities and cost are far beyond everything we have now.

DSC_0423-1452155492323.png
 

Chukzombi

Millie's Staff Member
71,639
212,860
It's not like they're gonna ban you bro. They'll just have "MEATBAG-ONLY" lanes where the speed limit is a slovenly 80mph and you have to watch people doing pilates or playing videogames at 150mph a couple lanes over.
we will see, currently these cars are only capable of 25mph because AI tech cant judge faster than that? and even if they could do 150mph its just going make the accidents more lethal. nope nope nopety nope nope. me and will smith will be the last ones still wearing chuck taylors and driving our old beaters to work. if i see a robot fucking my shit up i'll just jack that mofo.
 

Tuco

I got Tuco'd!
<Gold Donor>
45,395
73,466
we will see, currently these cars are only capable of 25mph because AI tech cant judge faster than that? and even if they could do 150mph its just going make the accidents more lethal. nope nope nopety nope nope. me and will smith will be the last ones still wearing chuck taylors and driving our old beaters to work. if i see a robot fucking my shit up i'll just jack that mofo.
An autonomous vehicle's max speed is dependent on the conditions and safety risks. Right now autonomous vehicles on highway can go pretty much whatever speed is legal. In many conditions the autonomous vehicle's perception is much better than a human.

In future systems the difference between an autonomous vehicle's perception and judgement and humans will be massive, but we're not there yet.


An annoying aside is that I doubt you'll be able to set the max speed of the autonomous vehicle higher than the speed limit. That wil lprobably be a fight between the OEMs and the government, because you'll essentially be giving the consumer the ability to tell a robot to behave illegally.

There's a project I'm hoping to work on soon that will involve trying to set a speed record on a race course with a fully autonomous system.