4 ms·

NYC is going to be an interesting case. As soon as pedestrians get comfortable that the cars won't hit them, I suspect a lot of jaywalking will happen - I wonde

NYC is going to be an interesting case. As soon as pedestrians get comfortable that the cars won't hit them, I suspect a lot of jaywalking will happen - I wonder if they'll get anywhere at all. You sort of have to br an aggressive driver too with taxis and delivery trucks, so I'm really curious to see how it all works out.

Animats 5y ago
You sort of have to be an aggressive driver too with taxis and delivery trucks, so I'm really curious to see how it all works out. There's a theory paper on this: "Go ahead, make my day: Robot conflict resolution by aggressive competition"[1] This is a known problem with mobile hospital robot carts. Many hospitals have robot carts moving linens and meals around a big plant. If they make no attempt to get others out of their way, they keep getting stalled by people talking in halls and crowding into elevators. So some degree of pushyness has to be programmed in. Cars, though. That may not work out as well. Maybe lighting and sound effects to encourage people to get out of the way. [1] http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.43.4978 http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.43.4...
- renewiltord 5y ago
  Great link. Thanks.
gen220 5y ago
Yep, I think this is how it’s going to work out, too. The dawdling driverless vehicles (paralyzed by opportunistic jaywalkers) will actually make the traffic worse. The worst possible future is if entrenched interests (waymo, etc) begin lobbying for more active measures of jaywalking enforcement (first policing, but also fences to keep those pesky humans on the sidewalk), as a way to get around the technical problem of becoming a more aggressive/communicative driver. The only natural conclusion of that possible future would be streets that are utterly hostile to humans, which would be throwing away one of NYC’s best assets. I hope it doesn’t come to fruition! And I hope driverless car tech can solve this problem. The idea is developed more / better articulated here: https://www.strongtowns.org/journal/2021/8/12/automated-vehicles https://www.strongtowns.org/journal/2021/8/12/automated-vehi...
- asdff 5y ago
  I really don't know why driverless car companies aren't going all in on busses. You can sell your tech to bus manufacturers, and if you lobby for BRT you have your own grade separated right of way which makes a lot of these monumental engineering challenges with self driving cars poof into thin air instantly. The biggest cost for a bus for a transit agency is the benefits package and salary for the driver.
  - danielmarkbruce 5y ago
    Because cars are more general purpose and thus significantly bigger market. Google is a $2 trillion market cap business. Building a $100 bill business, for example, isn't going to move the needle. It sounds crazy, but they sort of have to go after things this large.
    - asdff 5y ago
      Seems like you could bring busses to market a lot sooner given that automatic BRT is so much easier and use what you learn there for your private car stack. Plus you'd be sitting pretty in terms of getting public contracts and dictating regulation, since lawmakers will be turning to you, the domestic automatic bus software company, to weigh in on what to do for self driving car policy.
      - danielmarkbruce 5y ago
        Yep, there are many businesses which you could get to market a lot faster than self driving cars.
schoolornot 5y ago
Jaywalking aside, I'm interested in seeing how these vehicles handle the green arrows + pedestrian GO signals that occur at the same time.
itisit 5y ago
Jaywalking is already constant and ubiquitous in NYC. Most people simply do not observe crosswalk lights and wait to cross a street (not talking about avenues). And distracted pedestrians on their phones have brought the danger to a whole new level over the last decade. As you suggest, some bad actors will take it even further and provocatively test the safety response of automated cars. Pedestrian deaths are on the rise in NYC, and I believe beta testing self-driving cars will only contribute to that trend.

Animats 5y ago

You sort of have to be an aggressive driver too with taxis and delivery trucks, so I'm really curious to see how it all works out. There's a theory paper on this: "Go ahead, make my day: Robot conflict resolution by aggressive competition"[1] This is a known problem with mobile hospital robot carts. Many hospitals have robot carts moving linens and meals around a big plant. If they make no attempt to get others out of their way, they keep getting stalled by people talking in halls and crowding into elevators. So some degree of pushyness has to be programmed in. Cars, though. That may not work out as well. Maybe lighting and sound effects to encourage people to get out of the way. [1] http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.43.4978 http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.43.4...

renewiltord 5y ago
Great link. Thanks.

gen220 5y ago

Yep, I think this is how it’s going to work out, too. The dawdling driverless vehicles (paralyzed by opportunistic jaywalkers) will actually make the traffic worse. The worst possible future is if entrenched interests (waymo, etc) begin lobbying for more active measures of jaywalking enforcement (first policing, but also fences to keep those pesky humans on the sidewalk), as a way to get around the technical problem of becoming a more aggressive/communicative driver. The only natural conclusion of that possible future would be streets that are utterly hostile to humans, which would be throwing away one of NYC’s best assets. I hope it doesn’t come to fruition! And I hope driverless car tech can solve this problem. The idea is developed more / better articulated here: https://www.strongtowns.org/journal/2021/8/12/automated-vehicles https://www.strongtowns.org/journal/2021/8/12/automated-vehi...

asdff 5y ago
I really don't know why driverless car companies aren't going all in on busses. You can sell your tech to bus manufacturers, and if you lobby for BRT you have your own grade separated right of way which makes a lot of these monumental engineering challenges with self driving cars poof into thin air instantly. The biggest cost for a bus for a transit agency is the benefits package and salary for the driver.
- danielmarkbruce 5y ago
  Because cars are more general purpose and thus significantly bigger market. Google is a $2 trillion market cap business. Building a $100 bill business, for example, isn't going to move the needle. It sounds crazy, but they sort of have to go after things this large.
  - asdff 5y ago
    Seems like you could bring busses to market a lot sooner given that automatic BRT is so much easier and use what you learn there for your private car stack. Plus you'd be sitting pretty in terms of getting public contracts and dictating regulation, since lawmakers will be turning to you, the domestic automatic bus software company, to weigh in on what to do for self driving car policy.
    - danielmarkbruce 5y ago
      Yep, there are many businesses which you could get to market a lot faster than self driving cars.

schoolornot 5y ago

Jaywalking aside, I'm interested in seeing how these vehicles handle the green arrows + pedestrian GO signals that occur at the same time.

itisit 5y ago

Jaywalking is already constant and ubiquitous in NYC. Most people simply do not observe crosswalk lights and wait to cross a street (not talking about avenues). And distracted pedestrians on their phones have brought the danger to a whole new level over the last decade. As you suggest, some bad actors will take it even further and provocatively test the safety response of automated cars. Pedestrian deaths are on the rise in NYC, and I believe beta testing self-driving cars will only contribute to that trend.