Le Chantier

Urban Information Networks

<This is an English rendering of a talk I gave on November 18, 2011 at the Paris 2030 Colloquium.>

As you might imagine, I will talk about urban information networks.  Such networks have existed since the beginnings of cities; indeed they are one of the basic reasons for the existence of citiy:  to enable the sharing of information among the inhabitants so that they may learn from one another.  This is the same principle that enables millions of termites to federate their little intelligences and thereby create such amazing structures.  Information networks are thus a basic element of life in a city.  I am not thinking here of social networks, which deal with personal information, but rather of information concerning the how the city works, both in demand and supply of services. 

Historically much of this information was hidden and circulated only among a small, spatially-limited group of people.  It was in forms that did not easily permit wider transmission, integration with other information or mathematical analysis.  But today more and more of such information is available in digital form that does enable easy transmission, integration, and analysis.  So now we can begin to ask questions about the immediate needs of individual inhabitants and about the problems they face in exploiting the city's services.  Equally we can ask questions about the real-time performance of the city's services, their immediate capacities and the operational problems that exist.

I will talk about three themes in this space:  number one, The Invisible Rendered Visisble, number two, Information for Resource Management, and number three, OpenData 2.0.

On my first theme, I have already indicated the role that digital systems play in making information more widely and immediately visible.  This is central to enabling the inhabitants to benefit easily from the services and resources of the city.  It is also central to enabling city managers to make the best use of the capacities and resources that are immediately available.  A recurring problem we encounter in urban system management is that decisions are made based on information that is minimal or incomplete or stale.

A common example is the sharing of water among multiple cities within a single water basin.  Local city or utility water managers generally have no current information about how much water is available upstream in the system and about how much water their neighbours are drawing from it.  As a result, they will sometimes withdraw too much water, resulting in environmental problems or they will draw less water than they could, resulting in local supply limits or, in exceptional cases, increased risk of flooding.

An example of how information enables the inhabitants to most effectively use the immediately available capacity of the total, multi-modal transportation system comes from our work with CalTrans in the San Francisco bay area.  Here inhabitants with smart mobile telephones can subscribe to a service that enables CalTrans to observe their journeys based on the GPS reading from the telephone.  From these observations CalTrans can determine the individual user's common journeys.  When the system sees the user beginning a familiar journey, for example commuting from home to the workplace, it looks at the multi-modal choices available to the traveller and the operational status of each of those systems along the required paths, and then makes a recommendation to the traveller for the optimal way to make this journey at this time.  The traveller thus makes the journey with the minimum delays and disruptions and the transportation systems' loads can be balanced.

On my second theme, I am thinking of a project in which I am slightly involved in Zurich, where the technical university, the ETH, has developed a set of principles based on One Planet Living (OPL).  OPL seeks to define and encourage the adoption of a set of resource consumptions that will enable 9 billion people to fairly share the resources - water, food, energy, land - of the one planet we have available.  For energy, this results in an annual consumption of 2,000 kW-hours per person, or an average of roughly 250 Watts per person.  This goal has been adopted for an eco-district development in Zurich.  Achieving such a goal requires that the individual is aware of his or her own energy consumption, just as modern cars give us feedback on fuel consumption, and also careful management of the balance between demand and various sources of electrical energy.

If we now think of individual consumptions in existing cities, how close is each individual to this goal?  The question in most cases is unanswerable, because such information is not available unless the city or the utility has installed smart meters to measure consumption of electricty and water. This is a basic principle from Management Science - if you cannot measure it, you cannot manage it.

Another example of waste resulting from a lack of information comes from the construction industry.  A large city such as New York has of the order of a million buildings.  Since a building - depending on its purpose and how it was built and maintained - has a lifetime of 50 to 100 years, it follows that on average in New York some 10-20,000 buildings are torn down and re-built every year.  Imagine how much that represents in thousands of tons of old materials that are carted away and dumped at some growing distance in a landfill or in the ocean and in new materials that have to be created and transported to the building site.  How many hundreds of thousands of truck journeys on already crowded streets!

For the 2012 Summer Olympics in east London a different approach was adopted.  When the London Olympic Committee issued a contract for the preparation of the site for the Olympic village, it contained a requirement that 90% of the materials in the existing buildings should be re-used.  The management of this work went to an American engineering consultancy, CH2M Hill.  CH2 hired 2,000 unemployed people in east London, taught them construction Health and Safety and how to deconstruct the existing buildings.  Bricks and stone were cleaned up and prepared for use in the new construction.  Steel was shipped to China - at a peak of 300 tons per day - to be recycled.  Wood and other combustible materials were converted to energy.

The result was cheaper - CH2 completed the project on time and well under budget - than dumping the old materials and buying new and resulted in 94% of the old materials being re-used as well as a further 4% being used to produce energy.  Why could not New York implement such a system?  Because today there is no information available on supply and demand of such materials and hence no market for them.  A lack of information again leads to waste.

On my third theme I point to the future of exploitation of urban systems information.  You are no doubt familiar with the OpenData Movement, which is growing rapdily in popularity.  OpenData 1.0, as I call it, begins with publishing documents - government reports including eventually management reports on urban systems operational performance.  This is hugely valuable in developing transparency between government and citizens, something that is greatly lacking today in most cities. 

But an equally valuable step will come when cities begin to expose their operational information in near real-time by allowing it to be federated into private services.  The city in fact is the largest producer of information about what is going on, but today that information remains trapped inside the many agencies.  By publishing it, as the government of Singapore is doing and others - such as Dublink in Dublin, Ireland - are following, that information is available to be tapped by the innovative capabilities of citizens and enterprises.  While I strongly commend CalTrans' leadership in the example mentioned above, why should travellers depend on government to provide multi-modal transportation guidance?  Why not allow innovators to compete for the most effective algorithms to optimise such systems?  There are certainly some urban systems that are critical and whose operation should be reserved to the local government, such as water management, but in many other areas private enterprise is likely to be a more powerful source of innovation than public agencies.  This is what I think of as OpenData 2.0.

I have written elsewhere of this approach - government as an agent for capturing and publishing real-time information about urban systems operations - as the Stupid City, a term that refers back to the Internet principle of putting intelligence around the periphery of the network, rather than in the center as was traditional in the telephone network.

So these are my three thoughts for today on the importance of the flows of operational information in cities and the enormous impacts that can be achieved when these flows become digital.  This is a new way of looking at how cities work, a new way of helping the inhabitants to get the best use of the city's services, and a new way of engaging citizens and enterprises in making the city a better place to live and work.

Trains of Thought

"Look up the trains [to Little Purlington], Watson," said Holmes.

"There is one at 5:20 from Liverpool Street," he replied.[1]

Our son, Mark, gave me a copy of Alan Lightman’s The Discoveries for Christmas. This book contains reproductions of twenty two scientific papers from the 20C together with a preface to each one explaining what it says and its significance. Two of them, not surprisingly are by Albert Einstein from his Annus Mirabilis. In the year 1905 the then obscure, young clerk, working in the Swiss patent office in Bern, published four diverse works in the Annalen der Physik, any one of which would have earned him a Nobel Prize. If you have ever wanted to see how science is made, these are excellent examples.

One of these is the original article on what became known as Special Relativity. I had first become aware of Special Relativity as a high school student – almost half way back to 1905 – in books by some of the great British astronomers and cosmologists, such as Arthur Eddington and James Jeans. It is a hard work out for the mind to accommodate such seemingly irrational ideas, as I was to discover again a few years later when I encountered quantum mechanics. And thinking back to the glory years of classical Physics it seems truly astonishing that Einstein, working essentially in isolation, would have had such purity of thought to arrive at such remarkable insights.

I came to grips with this work originally through his thought experiment and only later through the mathematics and I have found such thought experiments enormously valuable in my own (much lesser) work.

At some point in the last few years I was struck by the thought that perhaps there was an association between trains – which feature so often in these thought experiments – and Special Relativity itself. Actually the thought was that the invention of the train timetable – presumably sometime in the mid-19C – is a rather unsung innovation of major proportions. The concept of a train timetable must itself have been quite a break through, since never before had there been a need to think about events that happened at precise times in remote places.

The need for train timetables arises not simply to enable passengers to know when they can catch a train or when they will arrive at a destination. Rather it is required to enable multiple trains to share portions of tracks safely by keeping them separated in time. Today of course this is all worked out by computers, but until quite late in the 20C it was a manual process based on drawing the movement of every train on large pieces of paper and noting where and when they needed to pass. No previous human had activity had required us to think of the coincidence of remote events.  The development of coordinated regional, national, and international train timetables seems to me to be a considerable intellectual achievement with no historical precedent.

Moreover, the printing of train timetables as pocket books - such as the Bradshaw’s Railway Guide employed by Sherlock Holmes and Dr. Watson - surely represents one of the first steps into the Information Age. The existence of such books is a first indication, perhaps only preceded by the concept of a map, of a world of information abstracted from the physical world and made portable and accessible. While the map and compass enable navigation in the spatial world, the train timetable and a watch allow navigation in the temporal world.

Einstein was certainly familiar with the well-developed Swiss railway system, but I have no way of knowing whether thinking about train timetables directly influenced him. Still, it struck me that it would have been far less likely for anyone to conduct the thought experiments underlying Special Relativity if the thinking underlying train timetables had not been already invented.

---

[1] The Adventure of the Retired Colourman, Sir Arthur Conan Doyle

La Forza del Destino

Lynn and I were in Montana during August for a short hiking holiday in Glacier National Park. It was a great experience, especially after the forest fire smoke cleared. We saw black bears, mountain goats, bighorn sheep, and a lot of alpine flowers. Back at the lodge I found an old copy of Michael Crichton's "Lost World", which is the sequel to "Jurassic Park" and great entertainment reading.

I thought I had read it already, but late in the book I came across a passage that took me by surprise. Like "Jurassic Park" the central theme of the book is how species emerge, evolve, and disappear. "Lost World" builds on this idea and introduces some ideas from Chaos Theory. The passage that grabbed my interest discusses some of the real scientific problems with Theory of Evolution.

To be really clear, this is NOT any kind of advocacy for Creationism in any form. The scientific community - that is those people who publish in peer-reviewed journals - have total confidence that Darwin’s theory is the true explanation of how historical and present day species emerged. Indeed we can now directly see evolution at work through genomics. But science depends on critical thinking and beyond that basic conviction of the soundness of the Theory of Evolution as the basis for life on Planet Earth, there are still some difficult problems to be solved.

The evolution of a species is driven by changes in the species’ DNA. If a change occurs that enables the species to better exploit an available ecological niche, then the species will probably be able to reproduce more successfully and, over time, to displace other members of its antecedent species that have not had the benefit of this evolution as well as other, unrelated species.

While this explains functionally what Darwin himself could not explain, it leads to a number of difficult problems. Large among these is the problem of the speed with which species have evolved. This problem arises because the modification of the species’ DNA is presumed to be a random process, for example due to errors in the biochemical processes of replication. At the level of these processes, it is impossible to determine what the macroscopic effect of a given change will be. The effect will only be manifest when the modified DNA is expressed in the creation of a new cell and ultimately an evolved animal. Only then will it be apparent whether the modification provides an advantage to the species. It is also quite possible that the modification of the DNA a) produces no effect whatsoever (because it occurs in a section of the genome that has no function), b) completely disrupts the replication (and hence rapidly disappears), or c) produces an effect that disadvantages the species (and hence disappears more slowly).

In effect we are throwing dice here with no idea of what the outcomes will be, but in a few cases we will get a double six and the species will be better adapted to its environment. In the realm of physics, this kind of behaviour is effectively described by thermodynamics. When a raindrop lands on a mountain top, every little movement of the water molecules that increases entropy (through downwards movement along the gravitational gradient) is preserved. An important difference between the random exploration of water molecules leading to the water running downhill and the random modification of DNA leading to the advantage of the species lies in the time scales. (Another is that life, unlike water molecules, has the ability to reduce entropy by creating increasing local order.) The water molecule’s Brownian motion means that it makes millions of experiments per second. The DNA modification requires multiple generations of the animal before we can say that it has succeeded.

We can estimate the rate of occurrence of random changes in DNA and we are also now beginning to understand which regions of the genome lead to the production of certain specialized cells, for example visual receptors. The problem we find here is that if we can only rely on random processes at the level of DNA to drive evolution, then it would appear to require far more time, perhaps 100 Billion years, for advanced life forms to emerge. Whereas we know that the Earth is less than 4 Billion years old and the universe only of the order of 15 Billion years.

So we have to question whether the process driving Evolution is simply random modifications of the DNA. This is dangerous territory. It is verging on the discredited philosophy of Teleology, which posits that the universe has a sense of direction, a programme that it is playing out. In “Lost World” the mathematician character suggests that the answer lies in Chaos Theory, which includes mechanisms that can produce very rapid changes in transitions between states.

The reason this passage caught my attention is that I started wondering many years ago about what I thought of as “the inevitability of the universe”. That is, that the physics (or mathematics) of this universe is so designed that it is inevitable that stable particles will form from an initial Big Bang, that nebulae will condense, that stars will form, and so on to the emergence of homo sapiens, just as it is inevitable that a drop of water landing on top of a mountain will find its way to the sea. This is of course close to the Anthropic Principle, but posits that designing universes may not be a random process itself.

I have no answers to these problems. They are members of the set of questions without answers, but nonetheless valuable in that they cause us to ponder deeply.