The Man Who Invented Intelligent Traffic Control a Century Too Early | Hacker News Reader

The Man Who Invented Intelligent Traffic Control a Century Too Early

spectrum.ieee.org

75 pointsboramalper39 comments

Now someone needs to invent a traffic light system that uses data collected by sensors and cameras to dynamically adjust traffic lights to improve traffic flow.

My brother and I envisioned this idea a decade ago. Imagine a system that can use deep learning techniques to automatically manage traffic to optimize travel times.

I wonder if this is desirable or if it's a case of predictability being better than efficiency.

It might be better to know it takes 30 minutes to get somewhere than to vary from 20 to 40 minutes depending on what mood the AI is in.

This has been done for quite some time, and it's unclear what value deep learning would add beyond the arithmetic we've used for a while.

https://en.wikipedia.org/wiki/Signal_timing

Major cities have Transit Operations Centers or similar where engineers watch road camera and sensor data (as well as local news and other sources of information) and direct traffic lights, gates, police, etc. to keep things moving. There was actually a controversy in London when this control room manipulated traffic lights so the International Olympic Committee officials evaluating London would always hit green lights, and perceive the traffic as better than it actually was.

This may be less apparent in some areas because it is actually done with the opposite intention: to slow things down. If you seem to be hitting every red light, it is probably because the local government decided to install a Traffic Calming program there in the interest of safety / fewer complaints from the neighbors.

If I was running for mayor of NYC I would focus my campaign heavily on cleaning up the traffic:

1) Negotiate with all those stores to move deliveries to night-time. So more parking opens up and they don't clog the streets. Any losses from eg tax breaks would be more than offset by the increased productivity.

2) Reward companies that allow employees to work from home one day a week, or that organize carpooling for employees. Install wifi in city subway CARS and run a campaign targeting companies letting them know their employees are more productive on the train with both hands free than stuck in traffic.

3) Get an official app for the city to let drivers know when parking spots open up. 30% of traffic in the city is cars circling around looking for parking. Charge a monthly fee for the app, and allow paying for parking via the app. However, I probably wouldn't want to allow people to renew their parking past 2 hours so the city could still make revenue from people who overstay. Also prevents people hogging the spots.

4) Apply the results of traffic research to tunnels and bridges which have heavy delays. Make all the traffic lights leading up to the tunnel reward 20mph driving, while at the tunnel exits it would reward 30mph driving. It's easy to force traffic to preset speeds like this using traffic lights, and you could do this for several blocks leading up to and away from the tunnels.

In short - less frustration, less pollution, less circling around, people get to where they want faster, make more money and city is better off.

Street parking is just a terrible experience all around. I'd prefer to see most street parking eliminated in favor of dedicated rights-of-way for bikes, buses, taxis, rideshare, etc.

If you want to have a car, you have to live somewhere that has an alley/garage space for you. If you want your business to be usable by drivers, have a parking lot or do a parking validation deal with a nearby garage.

I think the SpotHero/ParkWhiz model is very cool, would love to see it turn into a real-time auction with very precise competitive demand-based pricing. "Parking is surging right now, consider taking UberPool for $x less."

I already don't drive anywhere in Chicago unless I know there's an adequately sized parking lot, or I've booked reasonably priced parking ahead of time.

Seeing ahead of time that there is no reasonably priced parking generally makes me take the CTA, whereas otherwise I might have circled for a while before breaking down and paying $25.

> If you want to have a car, you have to live somewhere that has an alley/garage space for you.

You're obviously not referring to NYC.

Why not? That's the case in the centres of many European cities.

NYC is different. Most parking is street parking, or a paid garage parking, which can be expensive.

> Most parking is street parking

That's a decision the city has made, and the current state of much street-space in NYC.

Copenhagen has made the opposite decision: space in the city centre is allocated to bicycle lanes and footpaths on many streets. I live on such a street. If I bought a car, I'd need to rent a parking space.

London has an intermediate approach: resident parking on the street requires an expensive permit.

> If you want to have a car, you have to live somewhere that has an alley/garage space for you. If you want your business to be usable by drivers, have a parking lot or do a parking validation deal with a nearby garage.

I've wondered what would happen if a city was built with a mandate that there must be a parking garage on every block, or every X blocks. You could even make them a public utility, an extension of the public road network.

They do have mandates like that. In the D.C. area, every new multi-story apartment building must have at 1.5 car spaces for every apartment apartment larger than a small studio.

So what's the parking like in DC?

In any case, that doesn't quite cover what I was thinking of. I'm curious what would happen if there was a major city with enough parking for everyone to find a space within a block of their destination, whether that's home, work, shopping, or whatever.

> I'd prefer to see most street parking eliminated in favor of dedicated rights-of-way for bikes, buses, taxis, rideshare, etc.

These things only work for some of the residents (predominately younger residents without families).

> I already don't drive anywhere in Chicago unless I know there's an adequately sized parking lot, or I've booked reasonably priced parking ahead of time.

And those parts of Chicago suffer economically for it. Much more than what free parking would cost the city.

It's not by accident that you can chart the growth of civilization by the roads.

Making street spaces free would actually make the problem worse: the price is already well below what I'm willing to pay, the problem is that I'm unlikely to find a street space at any price (below the cost of a ticket) near a given destination.

To provide free parking in popular destinations such that it never completely fills up, the city would have to buy property (shutting down or relocating the businesses already on it) and build parking structures on it.

> These things only work for some of the residents (predominately younger residents without families).

Leaving aside the fact that seeing families on buses is not remotely uncommon, the same statement could be said of automobile modes. Those who can't legally drive (which probably constitutes a quarter of the US population) along with those too poor to afford automobiles are at least partially excluded from the benefits of automobile infrastructure. Given current technology, a transportation system for everyone needs to facilitate the use of a variety of modes, even if no particular mode serves everyone.

> And those parts of Chicago suffer economically for it.

Do you have any data for that? Most of the literature suggests that the provision of free parking reduces economic prosperity. The simple theoretical explanation is that parking should be provided at cost, just like most goods and services. But there is also empirical evidence to back this up. For instance when Old Pasadena raised street parking rates in the early nineties to deal with a dearth of parking in their commercial district: "The cities sales tax revenue tripled within the first year."* That's one anecdote, but consider googling Donald Shoup for more literature and data.

There are two additional theoretical benefits to using pricing to adequately supply parking. First, it favors persons who value their time relatively more than their money, which are exactly the kind of people who tend to buy stuff in stores. And two, increased parking rates can increase turnover, which in turn increases total sales. For more info on the importance of turnover, check out this SNL sketch: http://www.nbc.com/saturday-night-live/video/the-olympia-res...

*Speck, J. (2012). Walkable city: How downtown can save America, one step at a time. New York: Farrar, Straus and Giroux. (see page 134 for the quote on Google books).

Families across New York City give birth to children, get them to school, do groceries on the back of its transit system.

As do people in Hong Kong and Tokyo and Singapore and Seoul and London and Berlin.

Come and visit Copenhagen. You can see people from 6 to 90 years old riding bicycles, and younger children being carried in box bikes etc.

Most cities are trying to remove the private car from streets. There are principles around induced demand [1] and traffic evaporation [2] that come into play.

The principle of induced demand is that by making it easier to drive, you encourage car use, creating more congestion. You are better off making it harder for cars to access the city centres. Where major routes prioritise modal equality where separate space is provided for walking, cycling, and driving. Only after those are catered for do you provide on street parking.

Traffic evaporation is a term to describe the way humans think about driving. Most people perceive traffic in a similar way to the way water works in rivers. If you reduce flow then you create massive tail backs. However humans do not behave that way. What happens is that if a route becomes too difficult/congested, people look at alternative routes or even change mode of travel.

Parking control (you know you can't park in that area unless you have a permit) has been used extensively in NYC to discourage car use. Conversion of parking space to cycle spaces is also key to encouraging modal shift [3]. Nottingham UK, is using the workplace parking levy to finance and built its tram system [5].

Use of filtered permeability [4] is a significant tool in still enabling access to roads for motor vehicles, but making it less direct for cars by closing off roads as a through road. Rising bollard systems can enable a city to provide efficient public bus services.

The key bit to realise is that with urban design the car is no longer considered a good form of transport, particularly as more people move into cities. It is exceptionally inefficient and requires an immense amount of space as well as storage space.

I suspect that most cities will move towards a hyper aggressive approach to keeping cars on the edge of cities using a combination of parking control, congestion charging, making available routes into the city less direct for private vehicles.

I also see technology being vital to bridge the boundary between urban and rural living. Current public transport models are based on providing regular services using peak hour vehicle requirements. Yes a 120 capacity double length bus at 8am is right, but at 9pm it could easily be an 8 seater bus.

(I really should have applied for the ycombinator future cities team ;) )

[1] https://en.wikipedia.org/wiki/Induced_demand [2] https://en.wikipedia.org/wiki/Disappearing_traffic [3] https://en.wikipedia.org/wiki/Modal_share [4] http://www.cycling-embassy.org.uk/dictionary/filtered-permea... [5] http://www.nottinghamcity.gov.uk/transport-parking-and-stree...

I find the exuberance around cycling from this primarily California audience fairly grating.

Cycling is great... in the handful of months the weather allows for it in northern states.

It's also frustrating that cities are looking to make driving worse, but not (generally) to build rail infrastructure to make transit better.

I was once enthusiastic about carless urban life, but after experiencing it for a while... all the things you mentioned are likely to push me out into the exurbs.

When London looked at the population growth and how it could "keep" the city moving, they needed to get 5% of the population cycling all year round. It is very much a nudge psychology.

Yeah, but the funny thing about the all red timing that I've seen is that people can beat it by going faster than they should, which always struck me as wrong.

I do exactly this. :/

> This has been done for quite some time, and it's unclear what value deep learning would add beyond the arithmetic we've used for a while.

I've seen a lot of these claims, and they're easily debunked by driving through a city. The usual case is a platoon of 5 cars cruising down an arterial. A solitary car pulls up on a side street, and the light turns read for the arterial just in time to stop the platoon for the solitary car. A sensible algorithm would let the platoon pass first.

There are endless scenarios of these, to the point where the light algorithm finds a local optima where it is completely out of phase with the traffic and causes maximum delay and minimum throughput.

A relatively small amount of effort going into cameras and an adaptive system that minimizes a cost function would save billions of dollars in gas, time, and unneeded road widening.

There are two scenarios that inevitably get me thinking about smarter traffic lights. The first, is what you described, and the second is when I notice myself and a few others are sitting at a stop light while the intersection remains unused for 30 seconds.

Santa Clara, CA trumps out "retime the traffic lights" every time they want to raise taxes. Biggest crock of B.S. I've seen in a ballot.

> A sensible algorithm would let the platoon pass first.

You're taking an example of a sub-optimal local solution and claiming that is proof that the global solution is sub-optimal. That does not follow.

It could easily be, taking your example, that the platoon of 5 cars must stop at some future light, and the delay for 1 car is the smallest delay available for those cars. Therefore, stopping for the 1 car reduces the platoon's total travel time. As the driver, it's easy to think that was dumb because you'll never see the 15-car wait that was avoided by stopping early for 1 car.

Essentially, you are claiming that a greedy algorithm for traffic would always find the globally optimal solution (which we know is not true).

Walter isn't arguing for a greedy algorithm. He's giving an example of a local inefficiency and a mitigation. Lots of local inefficiencies add up to a global inefficiency. Illustrating a global inefficiency in itself isn't feasible in an anecdote. Thus I think your argument by assertion isn't possible to refute: you can bring up "local inefficiencies don't necessarily mean global inefficiencies" to dismiss all arguments that the current system is less than optimal.

Me, I've been stuck at traffic lights for 5 minutes at 2 in the morning, with no traffic whatsoever around. A very local algorithm would be highly preferable at that time of night, assuming usual (i.e. very little) traffic at that hour. It's an extreme example, yes: but there does exist a range of traffic patterns between the small hours of the morning and rush hour, and the same amount of global optimization effort is not applied to all hours, for obvious reasons. Something that could analyse the traffic using cameras and sensors and optimize wouldn't have to be paid extra to work late at night.

Several cities take the even simpler approach of having lights that change from timed cycles to blinking red/yellow after a certain hour.

Cars stopping and then starting again costs time, money, fuel, wear & tear, pollution, etc. It's far from an efficient solution.

Might not apply where you are, but in my state, if you sit at a red light for that length of time, with no traffic coming the other way, it is legal to ignore the light and go anyway - basically treating the light like a four-way stop sign.

Anecdotal but London seems to have a reasonably 'smart' traffic light system. It's something you only really notice when you're looking out for it (sign of a good system?), but as a pedestrian it's very often that pedestrian crossing phases match extended gaps in traffic flow. On roads with a good sightline, I often notice a gap in the flow ~30 seconds away that the lights will change in sync with.

The system is called SCOOT if you are interested.

I think traffic lights are ripe for disruption.

You would think. However, even the current UK systems are nowhere near optimal. Crossings that block traffic for too long; Cycle crossings that stop for longer than pedestrian crossings; Lights that are at red even when there are no other cars. If councils are not motivated to optimize the current systems, they are unlikely to be interested in anything new and disruptive? There is also the pain of taking any new ideas through the approvals system, which I can only guess is both not easy and will be gamed by the larger companies.

Outside my house is a crossing that waits until a gap appears in traffic before letting the pedestrian cross. Usually, by the time this happens I have already crossed the road (because a gap has appeared!).

Optimising "traffic flow" only seems to cause an obstruction for local pedestrians and cyclists. The only real purpose of a crossing is to restrict pedestrians so that excessive vehicle traffic can be accomadated. They should either prioritise pedestrians or reduce traffic.

You mean https://en.wikipedia.org/wiki/Adaptive_traffic_control, which we've used in every major jurisdiction of Australia starting in the 1970s?

In the UK too. Here's an interesting look at how vehicle detection works

https://www.linkedin.com/pulse/20141005200312-83983163-traff...

https://xkcd.com/277/

> “Because of my low grades, my father thought that I might be mentally retarded,” Adler later wrote. His father took Charles to see a psychologist, who said, “You have a remarkable son [with] a vivid imagination. He should be encouraged to invent.”

What a psychologist!

My psychologists said the same, as a child. I suspect it's a stock phrase for many of them.

Psychologists as a profession are much more comfortable with the idea of people whose thought processes are unusual, but still effective. It's what they're for, really.

Reply on news.ycombinator.com