Sign up for our newsletter
Community / Public space

“Reality mining” can help transform our cities. Here’s how

It’s Wednesday night, and Matthew is driving home on the outskirts of London after a meeting. The road is slippery from the rain, but he knows the route well and is driving carefully.

Suddenly a car edges into the middle of the road around the approaching bend, forcing Matthew to swerve as it passes. He attempts to correct, but the tyres lose their grip and the car crashes into the hedgerow, without anyone seeing the accident.  

Thankfully, the software in the iPhone on his dashboard has been monitoring the pattern of forces acting on the car, and, comparing it with known patterns of car behaviour, determines that it is very likely Matthew has had an accident. It automatically alerts emergency services and activates the speakerphone so they can communicate with him. They arrive speedily before he comes to any serious harm.  

This is just one example of an approach called reality mining that could change almost every aspect of daily life for the better. Through our devices, we already collect data for all kinds of purposes, but the truth is that we’ve only scratched the surface of the potential benefits. Reality mining aims to extend the practical uses of this data, by connecting real-life objects with advanced pattern-recognition capabilities.   

White papers from our partners

Road safety is one instance of how such techniques can be applied on a large scale in modern cities. Over a million people throughout the world are killed in car accidents each year: if problems could be identified before an accident occurs, lives could be saved. For instance, patterns of falling asleep at the wheel can be detected, and the driver alerted. In more serious cases – such as driving under the influence of alcohol – the car could be brought to a halt in a non-invasive and safe way.


Similar ideas have been around since the 1990s and exist in a limited way in some high-end cars, but require sensors to be fitted throughout the vehicle. However, due to the presence of gyroscopes in smartphones, we can now use an everyday phone as the sensor for most of these functions. This difference puts the service within reach of everyone.  

Reality mining innovations can be applied in almost any context. At a commercial level, they help financiers and insurance companies to model risk more accurately. They also allow customers to provide a further measure of assurance, leading to potentially lower premiums. Already, the model of insurance is changing from reimbursing against losses, to preventing losses before they occur, using this type of data-gathering framework.

In the field of health, reality mining technology combined with NLP algorithms can now detect pre-diabetes with close to 90 per cent accuracy. High-risk individuals can be attended to in detail through their credit card bills, Twitter stream, and phone activity. Patterns of fast food consumption and other risk factors can be fed back, so they can course-correct.

We can also look forward to a much wider use of data in the future. Transport for London already gathers data from Contactless and Oyster card usage, making it available (in an anonymised form) to third-party transport apps – so far generating time savings of £58m for customers. In America, 41 of the 50 largest cities use data-driven analysis to prioritise law enforcement resources, causing a drop in violent crime, property crime and murder of between 5 per cent and 15 per cent wherever the approach is used.

As cities become larger, they will become even more connected. Phones, appliance signatures and social networks will act as living sensors of a city’s social life, transport flows and maintenance needs. The new

mayor can encourage these developments by actively promoting London’s future as a Smart City. If we can creatively connect our cities in this way, we’ll learn how to improve our everyday lives immeasurably. 

Stephen Kong is a CEO of Think Eco, an “Internet of Things” company from British Columbia.
This article is from the CityMetric archive: some formatting and images may not be present.