How does an autonomous vehicle circumnavigate a closed road?

[Ruptga]

Very near future? HAH! Who is going to pay for that? I can walk out of my door and see a 20 year old car on the street right now. It will take 20+ years to get all the cars changed ONCE they become viable (which is another 10+ years).

Overvolt is right. The current Google self driving cars are road boulders. I get stuck behind them on a daily basis, I wish they would ban them from the streets (street has a 35 MPH limit, but they can only go 25. FU Google). They do nothing but cause traffic congestion.

Remember 15 years ago when the Prius was a big deal? People like you said the same thing about hybrids, and you weren't entirely wrong, but look at how many hybrids are around now. The convenience of a self-driving car is much greater than a car that only needs to be filled up 60% as often, so they'll probably be adopted even faster than hybrids were. In ten years one to two fifths of the cars on the road will have an autonomous mode, just you watch.

The current iteration is speed limited by law and liability, not so much by technical issues. Take it up with whoever is making the rules that apply to those vehicles, not Google.

 

[quikah]

Remember 15 years ago when the Prius was a big deal? People like you said the same thing about hybrids, and you weren't entirely wrong, but look at how many hybrids are around now. The convenience of a self-driving car is much greater than a car that only needs to be filled up 60% as often, so they'll probably be adopted even faster than hybrids were. In ten years one to two fifths of the cars on the road will have an autonomous mode, just you watch.

The current iteration is speed limited by law and liability, not so much by technical issues. Take it up with whoever is making the rules that apply to those vehicles, not Google.

The argument is that in the near future all cars will be automated, not 1/5 - 2/5. Not happening. How wrong were those people that thought this about hybrids? The highest marketshare in the US was 3.19% in 2013.

Google chose to expand their self driving car fleet using the Neighborhood Electric Vehicle (NEV)standard. Think golf carts for gated communities, or the small electric vehicles security guards ride around in. NEVs are generally limited to 25 MPH. The first iteration was built using Toyota 4runner hybrids (or some other Hybrid SUV, forget) that can drive normal speeds (still see these occasionally). Why did they choose to use NEVs for the expansion of the program? I have no idea. They don't belong on the roads IMO, but god forbid anyone questions Google...

 

[Red Squirrel]

My guess is they will have to start using larger area construction signs, or walls instead of just cones and word signs that rely on people reading the words. (ex: road closed, so you know not to go there, but a typical self driving car's radar might think it's just another vehicle and wait behind it)

I still say the biggest issue is going to be snow banks, they are irregular shapes that will be hard for the radar systems to interpret properly and it will be hard to tell what lane the car is even in. Can't use line sensors either because those are typically not visible in winter.

 

[Ruptga]

The argument is that in the near future all cars will be automated, not 1/5 - 2/5. Not happening. How wrong were those people that thought this about hybrids? The highest marketshare in the US was 3.19% in 2013.

Google chose to expand their self driving car fleet using the Neighborhood Electric Vehicle (NEV)standard. Think golf carts for gated communities, or the small electric vehicles security guards ride around in. NEVs are generally limited to 25 MPH. The first iteration was built using Toyota 4runner hybrids (or some other Hybrid SUV, forget) that can drive normal speeds (still see these occasionally). Why did they choose to use NEVs for the expansion of the program? I have no idea. They don't belong on the roads IMO, but god forbid anyone questions Google...

Sure, the argument you were responding to was about all cars in urban and commuter areas, and guess what areas have the most pressure for adoption? It's going to happen, and it's not going to take fifty years. Once it hits the market for real, 10 years for a noticeable chunk of cars on the road and maybe 30 for true mass adoption. Feel free to bump this thread if I'm wrong.

I'm pretty sure Google went after oversized golf carts because that's what was most practical given the regulatory environment. I don't entirely know though, I'm more interested in what Tesla is doing. I'm also pretty sure you're a bigger Google hater thanI am a Google fanboy, get over yourself.

 

[freeskier93]

My guess is they will have to start using larger area construction signs, or walls instead of just cones and word signs that rely on people reading the words. (ex: road closed, so you know not to go there, but a typical self driving car's radar might think it's just another vehicle and wait behind it)

I still say the biggest issue is going to be snow banks, they are irregular shapes that will be hard for the radar systems to interpret properly and it will be hard to tell what lane the car is even in. Can't use line sensors either because those are typically not visible in winter.

Relatively speaking, it's not hard to add a regular camera to recognize and read road signs, could possibly even follow detour signs.