The promise and possibilities of smart cities
Smart city decision makers will have access to oodles of data to base their decisions on, and better yet
decision-making will be made even easier as all options will be presented having been thoroughly analysed by sophisticated systems.
The promise of smart cities of the future is enormous — congestion, pollution, overcrowded transit, wasted energy, delayed emergency response all problems of the past. With smart cities, and I shall use this term as moniker that sums up the digitisation of analogue processes, big data, the internet of things, automation, machine learning, neural nets, artificial intelligence — almost anything seems to be within the realm of possibility.
Smart city decision makers will have access to oodles of data to base their decisions on, and better yet decision-making will be made even easier as all options will be presented having been thoroughly analysed by sophisticated systems.
But do we know if these sophisticated systems are really giving us the best option? I am astonished by the inability to often understand how AI come up with their recommendations or actions at times. When AlphaGo beat Lee Sodol, the machine developed a strategy that baffled the world of Go, a game that humans played for over two millennia. We can analyse and try to understand but how it arrived at it its conclusion is not clear. Recently, in one of our ongoing research projects, I asked the creators of the AI how it learnt to read a satellite images and create mapping at scale? The response was “we don’t really know [what] happens inside”. What we do know is that the AI is able to take our training material and create data at a scale, accuracy and speed that is unrivalled by humans.
AIs are very good at taking rules and playing by them. Clearly defined problems, that can be packed into an algorithm, will yield results at greater speed and higher levels of accuracy than can be produced by humans. Most importantly this releases us from trivial and mundane tasks to turn our hands to something potentially more interesting and meaningful.
Herein however lies the machine’s (current) biggest weakness when we think of the promise of smart cities. Cities are inherently messy, with the rules changing all the time. Messy as they all differ from one another in climate, politics, economics, social norms, size and so on. Messy as they are constantly changing and evolving, over the span of a day, a week, a decade, often in patterns we can only hope to recognise in hindsight. And most importantly, messy as they are made up hundreds of thousands, if not millions of humans, and all with their own individual quirks, personalities, moods and irrationalities.
When it now comes to making decisions over what is the best trajectory any one city should follow, we must recognise that we cannot possibly feed a machine all possible eventualities that will allow it to make the right choice.
For example, we often discuss the value of green spaces and vegetation in the cities of the Gulf. On one hand there is no way that it can be sustainable to grow grass and trees in the desert based on the amount of resources required to ensure their survival in the harsh climatic conditions.
On the other hand, how do we measure the joys of people using these spaces with their friends and families? Whilst we know of the benefits of biophilia, how do we place a value on it? I use a rule of thumb, asking myself if I would let one of my children run around without supervision on a piece of grass, then the effort expanded is OK. Consequently, on a landscaped highway junction, it is not.
We humans have 6,000 years of experience in making cities, and we must build on this. Modern city-making professions have their origins in mass urbanisation that the north-Atlantic world experienced during the first industrial revolution and the resulting squalour of cities in the 19th century.
Modern architecture, public health professions etc all stem from a desire to improve the cities for its residents. The last 100 or so years we have sought salvation in designing cities for the car. Now we are turning to another technology to help us resolve the problems it caused.