This youtube video explains how self driving cars will vastly improve traffic.
Imagine there is no more traffic jam on 101. How would that impact real estate? If you can reliably get to Mountain View or Palo Alto in 30 min from San Jose, it will help San Jose housing price tremendously. Being stuck in traffic is way worse than the sheer amount of time driving. It just frustrates people.
Also the reliability of traffic time is a big deal. It’s no good if I can’t reliably predict commute time within say 15 minutes.
Theory is great. .Practice is another matter…I doubt self driving cars will take over in my lifetime…But who kmows…surprise me…But I am a reactionary engineer…I only buy things that have been throughly tested…Not interested in being a beta crash test dummy…
Become a vegan, live a long life and it will happen in your lifetime. 
The ring example they give in that video is really flawed, because the net throughput of the ring stays the same, with or without the ripple. Not sure what that was trying to show. People think of traffic from their personal perspective “i.e., I’m stopping/starting, therefore traffic is bad for me”. What is actually important is the measure of throughput (cars/sec). It is not clear to me if ripples actually drop throughput.
Fundamentally, big traffic problems are caused by high throughput roads necking down to lower throughput roads. Highway 85 is a great example. That is the core problem of traffic. Once a core artery bottlenecks, it backs up into all sorts of connected feeders. Smarter cars might get slightly higher utilization out of these core arteries, but it is not going to be a step function
Drivers aren’t the problem, cars are the problem. We need to get cars off the road. We need public transportation, tax incentives for telecommuting, etc.
I would rather die young than be a vegan…In fact when the apocalypse comes, I will eat vegans first…I want a healthy source of protein. …lol…Oops I forgot…I am too old to die young…
How can the throughput be the same? If the average speed of a car decreases (that’s what braking does), that automatically implies the throughput decreases.
Throughput means you pick a point, any point, and count how many cars pass thru in a minute. If cars have to brake and later speed up again throughput will go down.
You are right, throughput goes down.
This type of “keep equal distance between cars” technology already exists. It’s in Tesla today. The software can also follow lanes and make turns. So it’s not science fiction.
But we need this technology deployed to Civics and Camrys not just cars costing 100K. I suspect even if only 20 or 30% of the cars have that technology highway traffic would already greatly improves.
Can we reach 20% penetration within 10 years? I think we can.
There are a lot of cars on the road that are 10+ years old. It’ll take at least a decade before self-driving is standard on all cars and not just luxury models. Then it’ll take another 10 years before a vast majority of cars on the road have the technology. So we’re looking at a minimum of 20 years.
Overrated… I prefer to live my life fully by taking all that life has to offer and when the time comes to check out, so be it.
The reason why I say this is because a family relative passed away recently. She was older (late 60s I believe) and was vegan and so was quite slim. More of an introvert too. Died. To me anyway, she did not enjoy life and here you are, dead, anyway. Sure, let’s not start smoking a few packs a day but come on moderation is the way to go in life. Seize the day, people!!! Seize the day!!!
Just wait for the next recession and a new Cash for Clunkers program (“Money for Modernization” ?)
The US government is broke. We’re not far from social security and medicare consuming every tax dollar collected.
There is a goldilocks throughput – that is the peak throughput the interconnect can support. The example they give goes immediately to the goldilocks throughput, where the cars are driving at the correct speed, with the correct distance in between (which is a function of speed, BTW).
The ripples are caused when the traffic requests a throughput greater than the goldilocks throughput (i.e., driving really fast and close together). The result is a push back, which drops throughput momentarily (everyone braking).
The net result is that in one system, the throughput is constant, in the other, the throughput is oscillating above and below the goldilocks, but both converge on the goldilocks throughput on average. That is the max the system can provide.
Regardless, none of this matters. These are very small effects. The net problem is having X amount of throughput feeding into an artery which only supports Y, and X >> Y. There are only two options here that significantly move the needle. Either increase Y (add more lanes), or reduce the number of cars on the road.
Drivers aren’t the problem, cars are the problem!
1 Like