| 1. | Brandon Donnelly | 14M |
| 2. | 0xdb8f...bcfd | 4.5M |
| 3. | jcandqc | 4.1M |
| 4. | 0x65de...c951 | 2.1M |
| 5. | kualta.eth | 869.1K |
| 6. | Ev Tchebotarev | 170.5K |
| 7. | stefan333 | 81.7K |
| 8. | voltron | 81.5K |
| 9. | William Mougayar's Blog | 28.4K |
| 10. | Empress Trash | 19.8K |
| 1. | Brandon Donnelly | 14M |
| 2. | 0xdb8f...bcfd | 4.5M |
| 3. | jcandqc | 4.1M |
| 4. | 0x65de...c951 | 2.1M |
| 5. | kualta.eth | 869.1K |
| 6. | Ev Tchebotarev | 170.5K |
| 7. | stefan333 | 81.7K |
| 8. | voltron | 81.5K |
| 9. | William Mougayar's Blog | 28.4K |
| 10. | Empress Trash | 19.8K |
I speculate a lot on this blog about what electric and autonomous vehicles will mean for the future of our cities. The reason it’s speculation is because it’s phenomenally difficult to know with any sort of certainty what the downstream effects of these technologies will be.
I’ve seen some people claim that a car is still a car. That is, all of the same rules will apply even if they’re powered completely by renewals and we manage to make drivers obsolete (5-10 years?). But I fundamentally disagree with this line of thinking. There will be both positive and negative consequences. They are just yet to be seen.
Benedict Evans recently wrote a post where he started to think about where some of these changes might happen. And so I thought it might be valuable to throw a few of these into the discussion mix. Here are some of his ideas:
About half of car maintenance spending in the US goes to things directly related to the internal combustion engine. Electric takes that away.
There are about 150,000 gas stations in the US. They go, along with their associated convenience stores, which is where the margins are made. Interestingly enough, more than half of all US tobacco sales happen at gas stations. Where does that go?
It is estimated that electric vehicles will increase overall electricity demand by 10-20%. But this could disappear with the battery storage and off-peak power.
Globally, about 1 million people die every year from car accidents. In the US, something like 90% of all accidents are thought to be caused by human error and about 1/3 of fatal accidents involve alcohol. Autonomy has the potential to take most of this away. Personally, I think we’ll look back and think about how dangerous driving used to be and wonder how/why we all did it.
A complete rethink of parking. This obviously gets talked about a lot. ~14% of LA’s land is thought to be used for parking. My guess is that parking ratios/requirements go way down (we’re already in the 0 to 0.3 per residential unit territory here in Toronto) and parking garages transform into yards for AVs.
Autonomous vehicles once again rewrite the retail real estate landscape. Benedict believes they will create more billionaires in real estate and retail than in tech or manufacturing. I like how he describes big box retailing as an arbitrage of land costs, transportation costs, and people’s willingness to drive and park. This point is likely about AVs + e-commerce. See yesterday’s post about Amazon.
Finally, his last point is that autonomous vehicles could become a kind of mobile Panopticon. The Panopticon was an institutional building typology conceived of by Jeremy Bentham in the late 18th century. It was based on the idea that inmates could all be monitored by a single watchman, without any of the inmates knowing if they were, in fact, being watched. It was a way of trying to impose strict obedience in prisons, and so on. Since virtually all autonomous vehicles require some sort of computer vision, Benedict argues that they could become the 21st century watchmen. Move over CCTV.
The other big question is about decentralization. New transportation technologies have consistently promoted greater suburbanization – think streetcar suburbs to car suburbs. The fact that you’ll be able to use your time more productively in an autonomous vehicle is continually floated as an argument for this trend to continue. But I haven’t made up my mind about this one.
Do you have any other thoughts on the downstream effects of electric and autonomous vehicles?
“If we have data, let’s look at data. If all we have are opinions, let’s go with mine.” – Jim Barksdale, former Netscape CEO
Fred Wilson wrote a post yesterday about Tesla’s data advantage in this self-driving car arms race that we are currently living through. (I found the above quote in the comment section of the post.)
In their Q3 2016 update, Tesla claims to have logged more than 1.3 billion miles on its vehicles equipped with Autopilot hardware. This is important because the more data it collects – across diverse road and weather conditions – the better the vehicles get at driving without human intervention. As Fred Wilson put it: “more data is better than more software engineers.” So that places Tesla ahead of Google, Uber, GM, et al.
I spent a lot of time driving over the past week, certainly more than usual, and I couldn’t help but think about how much better it would have been to instead sit in the backseat and read a book (or mindlessly scroll through Instagram).
I always try and use cruise control on long drives, but unless the road is fairly empty, I find it doesn’t work very well. Everyone is driving at different speeds and so I usually end up having to reset it / adjust it every so often.
The big question in my mind is still: How does the world look when driving longer distances doesn’t suck so much? What changes when you can get into your / a car (important distinction) at bedtime, fall asleep, and then wake up in a new place?
A lot, I think.
If you don’t follow the work of MIT’s Senseable City Lab, I highly recommend that you start.
Earlier this year, researchers from the Massachusetts Institute of Technology, the Swiss Institute of Technology, and the Italian National Research Council developed something that they call “slot-based intersections.” In a world where cars have sensors and drive themselves, it is intended as a more efficient alternative to traditional intersections. Goodbye traffic lights.
Much like air-traffic control, the way the system works is by assigning individualized time slots to each car for when they may enter an intersection. For example, in the diagram below (Sequence 01) the car approaching from the bottom left (#10) has a “stop distance slot” in front of it reserved for 3 of the cars that are currently in the intersection. The two that are traveling perpendicular to it and the car currently turning left into the same lane as #10 (on the other side of the intersection). The car in the midst of turning right (#5) is exempt because there’s no possibility of collision.
In Sequence 02 (below) you can see that car #10 is now turning left, which means it has its own time slot in the intersection. Other approaching cars now have a “stop distance slot” dependent on car #10.
In all cases, cars making a right turn are able to move freely, provided they will not interfere with any other cars.
The researchers estimate that real-time slot allocation might double the number of vehicles that a traditional traffic-light intersection can handle today and, in some cases, it might completely eliminate stop and go traffic.
Often when I write about self-driving vehicles I hear people tell me that cars are still cars. It doesn’t matter whether they are self-driving or not. The same inefficiencies apply. They are not the solution to urban gridlock. Elon Musk was also criticized (following his Master Plan) for not properly understanding urban geography.
But self-driving cars will create new efficiencies. I am not saying that they are a silver bullet, but I am saying that they will help a great deal. I don’t think that anyone truly understands the extent of these efficiencies, but there are a myriad of possibilities. This Senseable City Lab project is a perfect example.
What I am grappling with right now is the relationship between self-driving vehicles and traditional forms of public transit. Until we get a handle on the efficiencies and overall impact, it’s hard to ascertain how these different forms of mobility will work together. My gut tells me that the lines are bound to get blurry and that self-driving “cars” will feel less and less like the cars we know today.
Below is a video that was published along with the research. If you can’t see it, click here.
[youtube https://www.youtube.com/watch?v=4CZc3erc_l4?rel=0&w=560&h=315]
I speculate a lot on this blog about what electric and autonomous vehicles will mean for the future of our cities. The reason it’s speculation is because it’s phenomenally difficult to know with any sort of certainty what the downstream effects of these technologies will be.
I’ve seen some people claim that a car is still a car. That is, all of the same rules will apply even if they’re powered completely by renewals and we manage to make drivers obsolete (5-10 years?). But I fundamentally disagree with this line of thinking. There will be both positive and negative consequences. They are just yet to be seen.
Benedict Evans recently wrote a post where he started to think about where some of these changes might happen. And so I thought it might be valuable to throw a few of these into the discussion mix. Here are some of his ideas:
About half of car maintenance spending in the US goes to things directly related to the internal combustion engine. Electric takes that away.
There are about 150,000 gas stations in the US. They go, along with their associated convenience stores, which is where the margins are made. Interestingly enough, more than half of all US tobacco sales happen at gas stations. Where does that go?
It is estimated that electric vehicles will increase overall electricity demand by 10-20%. But this could disappear with the battery storage and off-peak power.
Globally, about 1 million people die every year from car accidents. In the US, something like 90% of all accidents are thought to be caused by human error and about 1/3 of fatal accidents involve alcohol. Autonomy has the potential to take most of this away. Personally, I think we’ll look back and think about how dangerous driving used to be and wonder how/why we all did it.
A complete rethink of parking. This obviously gets talked about a lot. ~14% of LA’s land is thought to be used for parking. My guess is that parking ratios/requirements go way down (we’re already in the 0 to 0.3 per residential unit territory here in Toronto) and parking garages transform into yards for AVs.
Autonomous vehicles once again rewrite the retail real estate landscape. Benedict believes they will create more billionaires in real estate and retail than in tech or manufacturing. I like how he describes big box retailing as an arbitrage of land costs, transportation costs, and people’s willingness to drive and park. This point is likely about AVs + e-commerce. See yesterday’s post about Amazon.
Finally, his last point is that autonomous vehicles could become a kind of mobile Panopticon. The Panopticon was an institutional building typology conceived of by Jeremy Bentham in the late 18th century. It was based on the idea that inmates could all be monitored by a single watchman, without any of the inmates knowing if they were, in fact, being watched. It was a way of trying to impose strict obedience in prisons, and so on. Since virtually all autonomous vehicles require some sort of computer vision, Benedict argues that they could become the 21st century watchmen. Move over CCTV.
The other big question is about decentralization. New transportation technologies have consistently promoted greater suburbanization – think streetcar suburbs to car suburbs. The fact that you’ll be able to use your time more productively in an autonomous vehicle is continually floated as an argument for this trend to continue. But I haven’t made up my mind about this one.
Do you have any other thoughts on the downstream effects of electric and autonomous vehicles?
“If we have data, let’s look at data. If all we have are opinions, let’s go with mine.” – Jim Barksdale, former Netscape CEO
Fred Wilson wrote a post yesterday about Tesla’s data advantage in this self-driving car arms race that we are currently living through. (I found the above quote in the comment section of the post.)
In their Q3 2016 update, Tesla claims to have logged more than 1.3 billion miles on its vehicles equipped with Autopilot hardware. This is important because the more data it collects – across diverse road and weather conditions – the better the vehicles get at driving without human intervention. As Fred Wilson put it: “more data is better than more software engineers.” So that places Tesla ahead of Google, Uber, GM, et al.
I spent a lot of time driving over the past week, certainly more than usual, and I couldn’t help but think about how much better it would have been to instead sit in the backseat and read a book (or mindlessly scroll through Instagram).
I always try and use cruise control on long drives, but unless the road is fairly empty, I find it doesn’t work very well. Everyone is driving at different speeds and so I usually end up having to reset it / adjust it every so often.
The big question in my mind is still: How does the world look when driving longer distances doesn’t suck so much? What changes when you can get into your / a car (important distinction) at bedtime, fall asleep, and then wake up in a new place?
A lot, I think.
If you don’t follow the work of MIT’s Senseable City Lab, I highly recommend that you start.
Earlier this year, researchers from the Massachusetts Institute of Technology, the Swiss Institute of Technology, and the Italian National Research Council developed something that they call “slot-based intersections.” In a world where cars have sensors and drive themselves, it is intended as a more efficient alternative to traditional intersections. Goodbye traffic lights.
Much like air-traffic control, the way the system works is by assigning individualized time slots to each car for when they may enter an intersection. For example, in the diagram below (Sequence 01) the car approaching from the bottom left (#10) has a “stop distance slot” in front of it reserved for 3 of the cars that are currently in the intersection. The two that are traveling perpendicular to it and the car currently turning left into the same lane as #10 (on the other side of the intersection). The car in the midst of turning right (#5) is exempt because there’s no possibility of collision.
In Sequence 02 (below) you can see that car #10 is now turning left, which means it has its own time slot in the intersection. Other approaching cars now have a “stop distance slot” dependent on car #10.
In all cases, cars making a right turn are able to move freely, provided they will not interfere with any other cars.
The researchers estimate that real-time slot allocation might double the number of vehicles that a traditional traffic-light intersection can handle today and, in some cases, it might completely eliminate stop and go traffic.
Often when I write about self-driving vehicles I hear people tell me that cars are still cars. It doesn’t matter whether they are self-driving or not. The same inefficiencies apply. They are not the solution to urban gridlock. Elon Musk was also criticized (following his Master Plan) for not properly understanding urban geography.
But self-driving cars will create new efficiencies. I am not saying that they are a silver bullet, but I am saying that they will help a great deal. I don’t think that anyone truly understands the extent of these efficiencies, but there are a myriad of possibilities. This Senseable City Lab project is a perfect example.
What I am grappling with right now is the relationship between self-driving vehicles and traditional forms of public transit. Until we get a handle on the efficiencies and overall impact, it’s hard to ascertain how these different forms of mobility will work together. My gut tells me that the lines are bound to get blurry and that self-driving “cars” will feel less and less like the cars we know today.
Below is a video that was published along with the research. If you can’t see it, click here.
[youtube https://www.youtube.com/watch?v=4CZc3erc_l4?rel=0&w=560&h=315]
Share Dialog
Share Dialog
Share Dialog
Share Dialog
Share Dialog
Share Dialog