Forecasting is a process of using historical data to make informed predictive estimates about the future. If used correctly, it can help us design more liveable sustainable, antifragile cities.
A recent report from the World Bank states that 70% of the greenhouse emissions are produced in cities and that ⅔ of the world’s energy is consumed in cities. More critically, we also know that the expansion of urban land consumption outpaces population growth by as much as 50%, which is expected to add 1.2 million km² of new urban built-up area to the world in the three decades.
Urban systems are complex. With these environmental challenges, cities and urban systems play a critical role in meeting our climate targets and improving people’s well-being and quality of life. Currently, there is a rising trend of urban sustainability. Urban sustainability refers to the revitalisation and transition of urban areas to enhance and improve liveability, innovation and reducing environmental impacts while maximising social and economic co-benefits.
The real estate industry can help in solving the challenges facing urban sustainability by focusing on:
– environmental (resource consumption with environmental impact)
– economical (resource use efficiency and economic return) and
– social aspects of sustainability (social well-being and health).
Urban sustainability is increasingly important, yet difficult to measure. The scope of it is broad, and it requires all sorts of measurements and indicators. So, how to take it into account while building and developing more liveable cities – places that are good to live in?
The growing trend of smart cities is trying to find ways to measure urban sustainability. Smart sustainable cities use data from various physical touchpoints to improve the quality of life, efficiency of urban operation and services, and ecological health, while ensuring that it meets the needs of present and future generations with respect to economic, social, environmental as well as cultural aspects.
But looking at data is not enough anymore. Developing cities in a truly smart, efficient way calls for integrating multiple sources of data and being aware of the trends of the future. This is when forecasting becomes important.
“For understanding forecasting it is sometimes helpful to draw an analogy with maps. Maps do not attempt to replicate reality with all its intricate details, constant updates and changes. Their purpose is to give an abstract, high-level representation of the territory and allow people to make informed pathfinding decisions.
In this regard, Forecasting models can be seen as maps for the future. They do not attempt to recreate the underlying processes with exceptional detail. Their main goal is to provide understanding about the big picture and the recurring patterns present in the data. In doing so, forecasts allow us to make informed decisions, alleviate risks and reduce uncertainty.”, says Tigran Khachatryan, data-scientist at CHAOS.
Hence, the power of forecasting lies not in the ability to predict every aspect of the future with meticulous details but rather in the ability to identify the significant trends despite all the details.
"Forecasting models can be seen as maps for the future. They do not attempt to recreate the underlying processes with exceptional detail. Their main goal is to provide understanding about the big picture and the recurring patterns present in the data."
Forecasts discover patterns in the data and tell us about the most likely outcomes out of all possible outcomes. This is what makes them the perfect tool for designing more robust, sustainable cities. Forecasts show us areas to focus our development efforts into, so that we can develop cities based on demands and maximise the usage of resources.
“When you want to create a livable city, you need to see the city as a complex adaptive system with numerous moving and interacting parts. Designing sustainable cities requires understanding of the demographic, mobility and area development patterns – all the pillars of urban sustainability. Forecasts can therefore give you a big edge in examining what is the best course of action to take in the long run”, Tigran says.
Forecasts also have a very unique way of speaking to us, that requires a sensitive eye from the one using them.
“If a forecast suggests that a population in a particular area is going to grow by 5% it does not mean that it is exactly what’s going to happen. It means that with all the available information 5% rise is the most likely outcome out of all other possible outcomes.”
Robust forecasts need to take into account the combined outcome of all available information. Sometimes it happens that people have information about a particular area. They might get surprised that the forecast is not in line with their expectations.
“This is a slightly flawed framework for looking at the forecasts. While it’s true that the people might have some insider information about a particular area, they cannot have all the information and know all of its potential effects. Nobody can have all the information. There can always be a piece of information that the person is not aware of that completely cancels out their expected outcome”, Tigran states.
Information that goes into the forecast must be validated, come from a trusted source and contain the combined outcome of all past events. Therefore, if we want robust forecasts, we should look at historical data. Historical data shows the combined outcome of all events that occured in the past.
There is a saying that you cannot drive a car by only looking at the rearview mirrors. This is especially true with forecasting. Although historical data is a very solid starting point, it’s not enough. Often there are external reasons for the observed behaviours. Identifying and using these external factors is part of what the data science team at CHAOS is constantly working on.
“Most forecasts assume that if something happens in a particular way, there are a finite number of reasons for it. Those reasons are predictors. Forecasts allow us to calculate how much of the variation in outcomes is due to the predictors and how much is just a random noise.”
Identifying the right external predictors and augmenting them with historical patterns is crucial and requires deep domain expertise. It has to be done carefully because not all correlation is causation.
“Without external predictors the models might not achieve the desired accuracy but include too many and it will cause an entirely different set of problems”, Tigran says.
In city development, you often need multiple sources of data. Using conventional and unconventional sources of data, combined with forecasting, provides a comprehensive understanding of our cities.
Moreover, sustainable cities of the future require holistically integrating multiple streams of urban data, both conventional and unconventional. CHAOS combines all these data streams and provides comprehensive tools for generating insights in one platform.
Furthermore, detecting patterns and anomalies help predict urban developments and how they will affect people. Our customers can step ahead with the right insights learned from the ocean of data.
Finding potential project sites is a key part of Lujatalo’s profitable business decisions. Now Lujatalo is supported in this by CHAOS, a solution offering location information data and urban forecasts. CHAOS has been implemented in nationwide use at Lujatalo.
CHAOS’ AI-based urban forecasts and data screening help noste.io promote a more sustainable real estate business and create value for its customers.
Multiple survey results and reports published during recent years indicate that the majority of executives in the RE industry now believe that harnessing the potential of big data and AI offers a huge competitive advantage But how far are we in the transformation process? Has the industry unlocked the full potential of data, advanced analytics, and AI?
