Sensors are changing how and where information is being gathered, creating novel information pools to tap into for better, more informed decision making.
It’d be a far cry to suggest that the story of human progress can be folded up into a neat, linear sequence. The less grandiose version of its history looks more like a patchwork of bold ideas that either grew in the face of resistance or perished as a result of it. A series of trials and error based off bold, but educated guesses and necessarily limited by the quality and quantity of information available at any given time. If knowledge is power in this scenario, then information provides the fuel.
Thanks to recent developments which have cut costs and increased efficiency in a variety of sensor technologies, we are able to embed them into a greater number of objects and devices. These sensors will passively gather and broadcast an immense amount of data about our lives and our surroundings — creating, in effect, a distributed intelligence — and empowering us to make some of the best decisions on earth, and we’re only beginning to understand and capture its true potential.
Martin G. Curley, a vice president at Intel Labs, explores how these connected networks are gathering data that will help inform the planning of future cities. In an interview, he explained what kind of innovative applications will be developed to engage locals and tourists alike, and how this distributed intelligence will help local governments make proactive decisions and increase their cities’ quality of life.
What’s the end goal of these distributed intelligence systems?
In cities — and it’s not rocket science — but the common vision that’s emerging is simply how can we make our cities a better place to live? How can we make cities run more efficiently and effectively? How can we make them more sustainable, through climate change mitigation, and so on.
We’re working to deploy fairly widespread sensor networks and integrate them into a sustainable intelligent system. Today, for example, in a London borough like Enfield, there are only three air-quality stations — and they know the air-quality can sometimes be poor — but by using the distributed Quark processor-based solutions that we’ve developed (instead of only three data feeds and a computer simulated model) they can get a hundred distributed different data samples for the same price, or less, than they’re paying now. With machine learning you’re able to constantly monitor and check the quality of the data, and make recommendations for mitigation actions and that can make a huge difference.
That in itself is useful. But when it’s applied as a part of a sustainable intelligence system? Suddenly it’s very useful. Be it air quality, be it traffic management, et cetera, this will be distributed intelligence in support of a goal. Moving from intelligence into real action.
Can you help us envision what some of the applications will look like?
Though the initial applications will be air-quality, sound monitoring, and then micro-climate and weather conditions — eventually, within each city, the different departments, such as environmental or traffic management, will be defining the specific characteristics of these collection parameters. Be it carbon dioxide, NOX or particulates — and it’s also effective to add video — because we know that air-quality and weather, or air-quality and traffic congestion, are co-interrelated, sensor networks start to become powerful when you are able to connect these dots.
As an example, an interesting project that we are doing in the district of Brixton, in the UK, is that we’re working with the schools. So today in London, 50% of all car journeys are less than three kilometers, and cars are the major source of pollution in London. In fact, the UK is facing a significant fine from the European commission — £300 million — because they haven’t addressed the problem. In addition, 20% of people in London get almost no exercise. The average person gets only 33% of the recommended daily exercise levels.
We’re integrating air-quality sensors and display boards outside of schools in Brixton to show parents and children what those short drives are costing them and their environment, and the children themselves are helping design the usage models for the application that supports this. By encouraging parents to walk their kids to school rather than drive, this is an example of where sensors are being integrated into solutions that are actually driving a behavior change and creating a lot of shared value. Improved air quality. Money saved on petrol costs. Kids and adults are getting more exercise and staying healthier.
How does real-time data play into the success of these systems?
Real-time data is important because it enables much faster responses. For example, one sensor that we’ve been starting to deploy is a noise-level sensor. If the noise level from a night club goes over a certain decibel level, cities may get phone calls from nearby residents. Then they have to respond and send somebody out to monitor the situation to see if it’s actually over the prescribed decibel level. If you can record that in real time, you will see when the allowed decibel level in a particular area has been exceeded, and deal with the problem before citizens start complaining.
More real-time sensing will allow for more automation which, ultimately, in cities, the goal is to run more efficiently and more effectively. You want to have a better quality of life, better service-level agreements for citizens and visitors. Automation is a great opportunity once you have distributed intelligence in place. As with the detection of unexpected sound. Because you’re already parsing the different sound patterns in a city you will immediately detect something like a collision. Right away we can give first responders a location and a timestamp enabling a much faster response than would be otherwise possible. The rules just need to be defined for specific parameters and automations. Then the system will gather for these and invariably give a good response — often better than a human would — and it will be cheaper, more efficient, more reliable. Essentially you’re taking variability out of the equation.
We’ve been conducting a pilot in Dublin with Trinity College and Dublin City Council where we’ve had about a hundred users using an app called City Watch. They can, for example, take a picture of an acute situation like a flood and tag it so that the city knows to respond. Likewise it could be a nice resource, like a green space, they’d like to tag and make available for somebody else in the city. Similarly, pedestrians could receive automatic recommendations for routes where the air-quality is better. If developers are able to add in elements of gamification and incentivization, they could build a virtual cycle where the more people add to it, the more they get back, and then it could pick up a lot of momentum.
Are there any concerns around privacy?
Of course, in the European Union privacy concerns and legislation are more stringent than in the US. I think the important thing is that citizens have the ability to elect in or out of what would amount to becoming a sensor on the network. Again, this is all new, it does become quite powerful when you have fixed sensing, mobile sensing, and then add what we call participatory sensing where citizens in the city are electing to report in real-time.
We’re in the early days of figuring out the governance of this big data — and that’s a very important consideration, and one which is evolving — but our approach is that we’re using living labs like the ones we have in Dublin, London and San Jose. We’re working with these cities, and learning together along with their citizens, about what will be the best way to govern, to curate and to act as broker for the data.
Our perspective at Intel Labs Europe is that we should strive to co-create this future with the citizens, and the city councils or local governments, of Dublin and London and so on. Whatever future exists will be something that has been shaped and created by a broad set of constituents. Niels Bohr once said, “Prediction is difficult, especially about the future.” That’s one school of thought, though we tend to share Allen Kay’s view which is that “The best way to predict the future is to invent it.”