Technology specialists are claiming that we are at the brink of a revolution for urban life. Real-time traffic adjustments, energy sustainability, efficient waste collection, and intelligent buildings are just a few of the many features of the so-called smart city. In this post, we will dive a little deeper into the technologies used by the city of the future and understand how mathematical optimization can leverage said technologies to enhance life quality.
A technology that always comes up when talking about smart cities is the Internet of Things (IoT). According to Kevin Ashton, the first person who coined the term, the basic idea is that currently, almost all data produced depends on humans who, by nature, have limited time, attention, and accuracy. However, if computers gathered and broadcasted data without human help, we could track everything. The goal is to empower computers with means of gathering information and remove dependency on humans. Yet, as we will see, collecting real-time data and making decisions that solve problems are two different things.
Having a city efficiently collect data without human intervention is a necessary step for making it smart, but the key component is decision making. All this information translates into value when used for better and faster decisions. To illustrate this, let us focus on the problem of energy efficiency. Many renewable energy sources such as wind and solar have a non-uniform production profile that is strongly affected by temporal cycles and weather conditions. To make matters worse, energy storage requires costly infrastructure, and energy transmission generates losses. Now imagine a city where residential and commercial buildings keep track of their energy consumption throughout the day, and weather conditions in the city surroundings are closely monitored. With that information we can use optimization models to support long-term decisions on investments in the infrastructure and automatically adjust short-term decisions based on online information.
Two technologies that deserve our attention as tools for making better decisions are machine learning and mathematical modeling. Machine learning is great for dealing with a large amount of data. It can be used to forecast variables, cluster information into meaningful indicators, and process data in a "human-like" manner. Mathematical models, on the other hand, are all about making optimal decisions. If you have well-defined objectives and a good understanding of your system, a mathematical model is the best option. Of course, in the complex challenges that surround urban planning, both technologies are needed together.
Analogously, industries are now investing in data collection with IoT and walking towards a future where things gather information and can make automated decisions. However, like in the case of cities, what makes an industry smart is the decision-making. Does your company have models of its systems? Do you know what the objectives are and how operational and strategic decisions affect them? Data alone does not automatically translate into good decisions; a deep understanding of the problem empowered by information is the real key to a smart future.
Author: Vinícius Mello - Consultant at Cassotis Consulting
Co-author: Fabio Silva - Senior Manager at Cassotis Consulting