For more than 100 years, IBM has been working to make the world a better place by enabling information to flow more easily and more rapidly. Twenty years ago the world began to build a global platform for sharing information. Today IBM is delivering an urban platform – the Intelligent Operations Center – to make cities safer, cleaner, less congested, less wasteful, and far more interactive and flexible. IBM Smarter Cities solutions demonstrate the power of information — the power to transform the industrial methods of centralized, siloed management and control into integrated, distributed and adaptive methods. Learn about the transformation in motion in both large and small cities around the world. IBM’s Intelligent Operations Center in Rio Video: IBM Helps Rio Become a Smarter City
- Dr. Colin Harrison, Distinguished Engineer and inventor of IBM’s Smarter Cities architecture (presentation)
We’re lucky to have Colin Harrison here because he and many of his colleagues who are in the room from IBM have been working very hard to build an integrated operation center solution, but I think Colin’s goal is to take us back a step from the product side of that and the particular service menu to thinking about the urban system of systems. We’re really pleased to have you here, Colin, thank you very much.
Thank you, Gordon. I’d like to thank Gordon and Jessie for inviting me here at this fabulous party ñ excuse me, conference, and for doing such a wonderful job on the organization of it.
What I want to do in these 15 minutes here is take you somewhat quickly through how we got into this mess. Where did these ideas come from? That’s fairly quick, but then close on where do I think it’s going now.
Let me start back around 2005. Am I going forwards or backwards here? Can we have some other slide? Maybe we don’t have other slides. Ah, now we do. Okay, alright, splendid here.
I guess in front of this audience it’s pretty much a truism to say that we live in a connected world, but back in 2005-2006, when our chairman started thinking about what he could see going on after we basically built a global network together, it was pretty new thought. He combined that with the notion that, not only are we very well connected globally and locally, but we’re also connected to a great deal of information that is mainly generated by transactions.
He could see in the future that we would have the compute power at affordable level to gain in more or less little time great insights from those huge flows of data. It’s what our friends at Google would call ‘looking at the data exhaust,’ where you look at the query stream to find out what is it that people are asking about. A large part of what we do in Smarter Cities is looking at the transaction flows, whether that is electricity charging or road tolling systems, to understand what is it people are trying to do in the city?
Then, starting at that time, around about 2006-2007, a key thought was how can we then make those processes more efficient. It’s just heard as there’re enormous gains that can be made in the operational cost of cities and the performance of the infrastructure in cities that we thought we could have a big impact through applying this idea.
Those flows of information mean that we actually know a huge amount, we can know a huge amount about what is going on today in the world, in a city, in a region in something like real time, that by applying analytics to the historical flows of that information we can find predictive indicators in it. We can actually know a great deal not only about the nowcast view ñ what is going on right this second ñ but also the forecast view of that. What is the forecast for congestion? What is the forecast for water demand? What is the weather forecast, in fact?
In many cases we’re already collecting that data. I have a lot of conversations about what sensors do we have to put in place in a Smart City to make it work, and my usual answer is, ‘Start by looking at the data you already collect,’ because many, many cities have this stuff piling up on disk drives ñ God bless them ñ but don’t really do anything with it. So that’s an interesting place to start here. Next slide.
This is a model that started to emerge around about 2007-2008 forward, of a physical world that we are now very much connected to through sensors, through local area networks of various kinds, that we could structure that data, we could extract and cleanse it, and put it into a consistent data model, and on that data model we could start to provide essentially a digital world, almost a mirror world of what was going on out there in the real world.
The interesting thing about a mirror world is you can do interesting things with it. You can run time backwards, you can run time forwards. You can run time faster than real time. So you can begin to experiment in that real world to understand what is going on, what will be going on, what could go on in the real world that you’re connected to. If you hit the enter key one more time.
This one actually comes from, I think, Christchurch in New Zealand, but you find this example all over the world today, where the bus knows where it is, has the GPS system on it. The bus can tell us what the GPS system is telling us. We can republish that information through channels like Open Data to tell the people waiting at the bus stop where the buses are. We can make predictions about when that bus is actually going to arrive at a particular bus stop.
That’s efficiency. That is our focus for the last several years, and still is today. I think if you hit the key one more time we have this sort of nice abstraction around instrumentation, interconnectedness and intelligence, the analytics behind this. That’s the sort of basic set of principles that the next slide shows us, I hope. Yes.
The portfolio of activities that IBM has developed over the last several years based around that set of thoughts, that the world is full of information. We’re all surrounded individually by a cloud of information. For those of you of a certain age, if you remember Pig-Pen, Charlie Brown’s friend, always surrounded by a cloud of dust. We’re all surrounded by this cloud of information, and out of that we can build intelligence systems.
We’re just talking about Rio ñ and I’m sure you’ll hear more about Rio later on ñ around public safety and leveraging the integration of information across a very large city to get a complete view. We like to call it the ‘single view of the truth’ that everyone can share to understand how now to make decisions about allocating resources for emergency management.
We’ve done a great deal of work around water, looking at predictive models there. This is part of Rio, as well, making predictions about where flooding is likely to occur. In Rotterdam, in the Netherlands, building models of the flow of rainstorms coming down the Rhine from Switzerland to Germany and France, and then colliding with storm surges coming off the North Sea ñ is that combination now going to overwhelm the dikes that are protecting the city of Rotterdam in the Netherlands in general?
A lot of work around traffic prediction, traffic information systems, as well. We can predict in Singapore up to an hour ahead with about 90% confidence that a given district in the city is in danger of becoming congested. That gives the traffic managers in Singapore then time to react to this, and they can do that through traffic information signs, by sending information to onboard navigation systems, and they can do something that, as far as I know, no other city in the world can do yet, which is they can in real time change the road usage pricing, change the tolling on the roads, and send a price signal out, which is one of the best ways of modifying human behavior.
I obviously could spend an hour or two wandering through each of these, and I would love to do that, and if we get a chance, I’ll be happy to sit down with you and talk about more of these. You can see it’s a very rich portfolio and it’s mainly built around this idea that by looking at what is going on in the real world, more or less in real time, harvesting insights out of those information flows, that we can help the city managers to achieve greater levels of efficiency, reducing operating costs, achieving theoretical capacities and things like transportation systems.
We kind of had got to the point ñ we haven’t built any of it by that time, but by about mid 2008 we had figured out that these were the kinds of things that you could focus on and the army went off and started building these things. But the real penny didn’t drop until probably August-September of 2008, when a bunch of us stumbled into a project you might have heard of outside Abu Dhabi called Masdar. Masdar had, a bit weakened today, but at that time a very visionary idea of being a self-sufficient city, self-sufficient in many things, particularly energy.
We looked at this problem and said, ‘Well, energy has to make water, it has to drive the cooling system, it has to drive the transportation system, it has to keep the lights on, it has to keep the factories running. What do you do, since this is all solar, what do you do when the sandstorm comes? How do you decide now to allocate resources across the many different consumers of electricity, as it was at the time?’
What we realized then was what you really are looking at here is a ñ you could call it a control system, we actually I think preferred to call it ‘a management system for a city’ that looks across the different agencies to see their interdependencies, and then how now do we make overall that city operate better, rather than just looking at it silo by silo.
Let’s go on. I would like to spend the remaining time on this one last slide. We went from efficiency in different silos to this notion of an integrated management system for a city to the future. The thoughts that are intriguing me these days are starting to – it’s like the internet all over again. Remember the internet 1990-ish? We could just about do email on it. I remember when AOL first connected to the internet, that platform became the basis for the early web. Early web drove us to higher bandwidths like video streaming. That became the basis then for gaming, and so on. We built layer by layer, obviously. Given the Smarter City infrastructures that are beginning to deploy, what else is that going to enable?
One of the things, the models that I’m thinking about these days is this notion that what we’ve created now is essentially an interface, a user interface between the city and the citizen. By that I don’t mean the gooey on your iPhone or on your PC. What I mean is that there is a flow that can be – flows of information – in both directions there. Think about it as a user interface problem or a pure design problem, and ask yourself, ìHow do you improve the usability of a city?î One aspect of usability is discoverability. How do I discover function? If I get a new version of Microsoft Office, how do I discover the new functions that Microsoft has put into that? And then how do I understand that they bring value to me? If I go to a new city I’ve never been to before how do I discover how to make a journey through that city? How do I discover how to change money or to find a meal? Conversely, how can the city discover what it is that I am trying to do, either individually or as part of a cohort of some kind? Because if the city can understand what a group of people is trying to do, it can do something quite original.
If you think of our traditional transportation system going back several thousand years, the way you manage that system is you publish a timetable, and if I want to take the boat from Southampton to New York it’s my job to be on the dock in Southampton just before that boat leaves. In other words, you make the passengers organize their lives around the service offering. But what we’re getting close to here is being able to invert that and now begin to organize a service offering around what the citizens are trying to do. This is quite a revolutionary thought it turns out for a lot of cities.
One city that we studied a lot is one of our living labs is the little city of Dubuque in Iowa, 60,000 people. It’s a city that, like a lot of cities that struggle these days, had lost its primary industry and is looking now to reinvent itself. We did origin destination mapping of a number of volunteers in that city and we discovered that the transportation system still reflects the way the city worked 40-50 years ago. But that isn’t where people live and that isn’t where people work today. How can we help that city now to reduce its transportation costs by 30% and reduce journey times substantially? Doing that for a city of 60,000 people is a great discovery; they’re very happy with that. Now we’re going to do that for Istanbul. That’s two orders of magnitude bigger. That’s a lot more interesting.
I think I’ll begin to wind up at this stage, mentioned this idea of the cloud of information that surrounds us. This is a deep thought by some of my colleagues in IBM around what we call sense making, which is ñ now I’ll come back to this idea of connected world. In a connected world I have lots of choices. That’s an aspect of usability, but that also brings with it a lot of dependencies. Dependencies connects me now not only to good things but to bad things, as we discovered in Japan last year. If a factory in Japan is out of operation because it lost its power, a car factory in Detroit may be shut down shortly thereafter. Out of this immense flow of information, how do I determine whether a given piece of information is important to me? In other words, how do I make sense of that immense flow of information?
That’s where I am on the future of Smarter Cities. I’m thinking about this now in terms of how do we exploit the inflows of information that we’re beginning to create here. I’ll stop at that point.
Thank you, Colin. Fantastic.