Autonomous Intersection Management: Traffic Control for the Future

Channel:csatc2002 · Length:3:56 · Published:2012-06-27 · Views:419353

Watch on YouTube

Summary

A short research demo of Autonomous Intersection Management, a protocol in which self-driving cars reserve slots through an intersection instead of obeying signals or stop signs. It uses UT Austin's autonomous vehicle Marvin and a simulation to show cars crossing without stopping.

Why HN recommends it

Cited as the canonical visualization of signal-free intersections whenever HN discusses autonomous cars and road throughput.

you can just have the traffic streams pass right through each other. — HN, 2016

Self-driving & autonomous vehicles · Transit, cities & urban planning

Subtitles (excerpt)

autonomous intersection management traffic control for the future a project from the University of Texas at Austin most intersections today are controlled by traffic signals stop signs or roundabouts however even when people do follow these protocols intersections are the source of much traffic congestion and many accidents even though they take up a small percentage of the roadways roughly one quar of all accidents and onethird of all fatal accidents occur at intersections it is now apparent that autonomous cars are well on their way to becoming a reality at first they will need to follow the same intersection control protocols that were designed for people but this project considers once most cars on the road are autonomous whether we can do better than that by way of an AI based coordination mechanism in this research we introduce a new multi-agent intersection control scheme specifically designed for autonomous vehicles by utilizing much more of the intersection at any given time than traffic signals and stop signs the aim reservation protocol has the potential to dramatically reduce congestion and therefore fuel usage by eliminating the need for a lot of deceleration and acceleration harmful emissions can be dramatically reduced as well aim relies on

On Hacker News

Traffic isn't a function of road space. The true throughput of roads is immense when utilized by better drivers (one extreme example: [VIDEO] ). A lot of traffic is caused by bad human drivers: accidents (and associated rubbernecking), shockwaves, running red lights, double parking, circling around looking for parking. Additionally entire additional lanes of traffic (for cars or personal mobility… — xnx, 2023-05
> I can't help but think about how much this would alleviate traffic congestion Especially given that I do not see how self driving cars could decrease traffic congestion problems. A self driving car takes as much space as a "regular car", so only ways I can imagine autonomous cars reducing traffic are: 1. People ride less with cars. 2. People share rides more. 3. Autonomous cars are significantl… — beefield, 2016-12
> (Well, maybe one more option came to my mind that might actually help. If the autonomous cars would be able to increase the capacity of crossroads by somehow communicating with each others like this: [VIDEO] ) How do cyclists and people walking handle that type of intersection? — chris_7, 2016-12
> Self-driving cars ... could change this equation by intelligently routing cars around bottlenecks and avoiding the neighborhood entirely Self-driving cars can do even better than that - they can eliminate the "stopping" nature of the intersection entirely. If the self-driving cars are able to determine the position and speed of other cars approaching the intersection, you can just have the traf… — coryfklein, 2016-06
It's very likely that once self-driving cars reach a certain penetration, we'll see continuous-flow intersections where no human driver would dare try and cross. [VIDEO] — omegaworks, 2015-10

Related videos