Are We Closer than We Think? Defining and Creating a Smart City with Modelling & Simulation

By Dan Macklin , CPTO, Skyral

The idea of Smart Cities isn’t new, but it certainly has blasted onto the scene, often compounding in conversations around how we are going to effectively govern in this new AI age. But even with the trend-spread of Smart Cities so quickly becoming part of our everyday vernacular, as with many technical concepts, the actual meaning is pretty amorphous. What is a Smart City? Where does this vision come from? And what is the roadmap for actually making the cities of today, and tomorrow, ‘smart’??

I’m an advocate for breaking things down into their most basic parts to help understand how concepts fit together. So let’s do that here. Let’s start with how we define a ‘city.’ No, not Merriam-Webster style, but from a technical perspective: a city is an interconnection of different systems.

By systems, I want you to think about interactions, because it's those interactions that make up the systems that make cities work. A road network that commuters use to drive to work is a system. Meanwhile, the bus route is also a system, and so is the train system, and the tram system, and so on. Each of these systems are deeply interconnected and their varying dynamics present a serious amount of complexity. Then add to each transport system the city’s power infrastructure system. And the water system. And the telecommunications system. And then add the cultural, economic, and societal norms of each individual in this system as they go about their normal daily lives.?

With this in mind, a Smart City is one that is able to adapt to the dynamism presented by these interconnected systems upon systems so that they run smoothly when stress and chaos are low, but respond and show resilience when unexpected events inevitably occur.?

The ‘utopian’ idea of Smart Cities seems utopian because the general hegemony assumes we aren’t able to predict - just react, and therefore can’t effectively respond to that inevitable chaos. “How?” many ask, “could we possibly understand these systems in both a steady state, during major system change (adding in a new rail system) and during chaotic events?”?

This won’t surprise you, but what we’re doing at Skyral gets at answering these types of questions, and the main reason I think that is, is because we’ve cracked one of the biggest blockers to understanding those systems: modelling and simulation at scale.?

Scale equals nuance. The reason we assume we haven’t cracked the code to developing Smart Cities is because the traditional modelling technology used to help optimise for the day-to-day activities in the city have a very real limit. They can only assume that everything is running as it should, and will run in that same way far, far into the future.?

Most modelling tasks, such as traffic management, focus on optimising systems for typical, everyday conditions. These models are designed to ensure smooth operations during typical days by predicting and improving performance based on average scenarios (an analogy might be weather modelling.... the average weather for a given day is easy to model and has value but detailed hour by hour regional forecasts require massive scale and huge datasets). This approach aligns with the central limit theorem, a fundamental principle in statistics, which states that as the number of data points in a model increases, the overall behaviour of the system tends to average out, making extreme or unusual events less likely.

However, the real world often throws unexpected challenges our way. These unpredictable, nonlinear events which cause significant disruptions are difficult to test using traditional models because they are built to handle what's typical, not what's chaotic and interdependent. Our customers are increasingly interested in understanding how these events happen and how small changes in one system can have major consequences in others.

This is where agent-based modelling (ABM) becomes crucial. Unlike traditional models (linear regression, systems of equations, discrete events, systems dynamics), ABM doesn't focus on the average outcome. Instead, it simulates the interactions of individual agents, whether they are drivers in traffic, consumers in a market, or any other entities each following its own set of complex behaviours. These interactions often lead to surprising and sometimes extreme outcomes that a typical model might miss. In ABM, the central limit theorem doesn't apply in the usual way because the system's behaviour is driven by these dynamic interactions, not just by averages.

What's particularly important is that the more agents you include in the model, the better you can capture the full range of potential interactions, including those that might lead to rare, outsized events. In smaller models, the diversity of interactions is limited, and the chances of observing these rare events are lower. But as models grow, so does the complexity and variety of possible outcomes, making it more likely to spot and understand these black swan events.

In simple terms, when it comes to agent-based modelling, "bigger is better." Larger models allow us to simulate a wider array of possibilities, giving us a better chance of predicting and preparing for those rare but impactful events. This insight can be a game-changer for our customers, helping them not just optimise for the typical day, but also prepare for the unexpected.

Now, I appreciate that got pretty technical, pretty fast. So let’s recap:

A city is a system that's able to react to chaos in a really dynamic way because of an inherent understanding of patterns and the systems that make it up. So it's a place (and this might sound really silly) that is proactively reactive. A Smart City is one that's able to react to those unexpected events, because the precursor to that event has already been proactively looked at. People are asking, everyday, ?“okay, what happens when these black swan events take place in a system that is already chaotic?” And from modelling these scenarios, they’ve created resilient solutions before the event happens in reality. A Smart City is one whereby you can make your mistakes in a virtual world and plan them out before they happen in the real world, so you're more prepared, more resilient, and more protected.?

Long story short, we are far closer than ever to the vision of the ‘Smart City.’ The capability from a technology perspective is there. But how do we get this off the ground? Organisations, and I mean both businesses and government organisations, must foster two main characteristics to make Smart Cities a reality.?

First, they must be curious. We need people who are curious about the different possibilities. As I said before, the difference between modelling the steady state and the unexpected state is that people who want to simulate the black swan events are more curious, and our technology can help power that curiosity, because we can build hundreds of different types of scenarios, hit a button, run them, and have some answers. And from those answers, users can view their solutions at a statistical level, but they can also rerun those simulations in real time, and then interact with potential solutions. That’s curiosity in action.?

Second, they need to be sceptical. You might have heard the term ‘human in the loop’ come up in discussions about AI taking over the world. At Skyral, we believe a healthy dose of scepticism helps keep technology on its toes, and keeps the human in the loop. To make Smart Cities a reality, users need to be sceptical about what machines are churning out to ensure that the black swan events we’re responding to are logically and appropriate to the context. A machine or a model can only take us so far. To get to that state of resilience, synergy, and efficiency requires people to flex with new technology and keep it in check.?

It’s easy to think of Smart Cities as flying cars and hydroponic farms on every rooftop. In reality, a Smart City is one that reflects the lives, attitudes, and daily traditions of its inhabitants, responds to chaos in the calmest, most effective way possible, and delivers a kind of resilience that enhances and protects all life in the system.?

We’re closer than we think.