One brain to rule the city

I’ve been watching YouTube videos of Alibaba’s City Brain in action, and I must say I’m impressed.

The City Brain is an attempt to take all the data generated in a city and to use that data to manage the city by making intelligent decisions in real time.

So, for example, cameras trained on a road intersection produce massive amounts of data, but it’s in a format that is very hard for people to use effectively. This data can be interpreted, using algorithms, to identify the type of traffic and the directions in which it is flowing. The City Brain can then change the timing of traffic lights to improve the traffic flow. It can also identify problems, such as a stationary vehicle blocking a lane, or an accident, and alert the appropriate authority.

Another example is connecting fire alarm systems with street maps and building plans. Instead of just sounding an alarm, a smoke detector can inform the city fire department exactly where a fire is, can work out a route for vehicles to get to the fire, and can control other traffic along the way to allow quick access. It can also give fire-fighters information about the building, how to access it and how to navigate around inside it.

The City Brain combines data from maps of the city, cameras, traffic lights, navigation systems in public and private vehicles, sensors around the city, social media apps and networks. It provides services that include detecting traffic events, predicting congestion, optimizing traffic lights, dispatching emergency services and adjusting bus frequencies based on demand. One of the benefits of the city brain is that it’s a software based solution that uses existing data sources, so its relatively low-cost. The system does not need new infrastructure installed to make it work.

The system was first implemented in Hangzhou, China where it is reported to have increased travel speeds, reduced travel time, reduced emergency response times, and increased the accuracy of identifying incidents. It has since been rolled out in four more cities in China as well as in Macau and Kuala Lumpur. There are plans to roll it out in the Middle East.

All this is cool, but the best part, for me, is that drivers in these cities report that they drive better because they know that they are being watched.

I get excited about this because I’ve tried, too often, to drive out of Braamfontein (in Johannesburg) at 5pm on a weekday. Some drivers in Braamfontein think nothing of swinging out onto the right-hand side of the road (we drive on the left) and facing down oncoming traffic if it gets them a few car-lengths ahead. Trying to drive across an intersection when errant vehicles on your right are crossing in front of you to turn left, is a heart-stopping experience. So I really like the vision of order that a City Brain might bring to urban traffic.

Effectively, the City Brain acts as a panopticon, keeping an all-seeing eye on city drivers that results in them policing their own behavior. Now the panopticon design was intended for restrictive institutions like prisons, factories, asylums and schools, all places which sought to control the behaviour of many with the oversight of a few. Is it a good thing to constrain people so that they feel obliged to drive carefully because they are being watched?

I think so. People are not good at self-regulating. They do stupid things that make life in the city dangerous, so an external impetus to behave better improves the city for everyone at minimal cost to the individual. The City Brain is not trying to limit freedom in the way that a prison (or school) does. Rather, it is trying to make the city a better place in which to live and get around.

Could the City Brain become a monster that tries to control individuals more closely? Yes, it could, potentially, decide that you should not go to a particular part of town and you might find all routes there blocked. So we would need to make sure that important human freedoms are protected.

Does the City Brain make use of personal information in intrusive ways? That’s not clear from any of the information I have seen. I would hope that the use of individual social media data is something that can be opted into or out of. The system ought to be able to work with aggregate data, for example to assess the demand for public transport. But I can also see the benefit of using personal data. For example it would be helpful if an ambulance could be informed that the person involved in a traffic accident is diabetic. Presumably each city can decide which data can be used in which situations, in line with relevant legislation.

I think that big cities are going to find it hard to resist this kind of management tool, if it does deliver the benefits it claims. So a conversation about the potential implications, unintended consequences and where we draw the line for data use, is probably worth starting.

Drivers in Hangzhou clearly feel that they can be identified, presumably through car registration information. When cars are driver-less, the close connection between an individual and a vehicle will go away. There will also be no need to control the behaviour of the driver. Presumably, driver-less cars will obey traffic rules in the first place and when the City Brain does need to control them, for example to take charge in an emergency, it can be done directly with no messy, conscious, fleshy human getting in the way.

Personally I’m hoping that the whole traffic problem will fade away as private ownership of cars becomes a thing of the past. Shared car services and public transport will be good for the planet, and also for the physical spaces we inhabit once we grass the parking lots and plant trees. I sold my car earlier this year and I hope never to have to own another. In the meanwhile I can’t wait to have the City Brain tackle the traffic in Johannesburg.

Image by Tumisu from Pixabay.