Sunday, March 4, 2012

The future of urban management - smart cities

The Times has a nice article that chronicles IBM's experiment in Rio di Janeiro to reshape the future of urban management,

City employees in white jumpsuits work quietly in front of a giant wall of screens — a sort of virtual Rio, rendered in real time. Video streams in from subway stations and major intersections. A sophisticated weather program predicts rainfall across the city. A map glows with the locations of car accidents, power failures and other problems.


As the Times article writes, it is increasingly possible to use powerful data analytics software to forecast trends and thereby provide decision-support on various issues of urban management. This is expected to help decision makers anticipate, instead of react, problems and plan accordingly to either avoid them or to atleast minimize the damage (or derive more benefit) from them.

Many metropolitan areas already use data-collection systems like sensors, video cameras and GPS devices. But advances in computing power and data analysis now make it possible for companies like IBM to collate all this data and, using computer algorithms, to identify patterns and trends.


IBM is the master integrator who co-ordinates the functions of all other partners,

Local companies handled construction and telecommunications. Cisco provided network infrastructure and the videoconferencing system that links the operations center to the mayor’s house. The digital screens are from Samsung. IBM coordinated everything... IBM incorporated its hardware, software, analytics and research. It created manuals so that the center’s employees could classify problems into four categories: events, incidents, emergencies and crises. A loud party, for instance, is an event. People beating up each other at a party is an incident. A party that becomes a riot is an emergency. If someone dies in the riot, it’s a crisis. The manuals also lay out step-by-step procedures for how departments should handle pressing situations like floods and rockslides...

IBM also installed a virtual operations platform that acts as a Web-based clearinghouse, integrating information that comes in via phone, radio, e-mail and text message. When city employees log on, they can enter information from, say, an accident scene, or see how many ambulances have been dispatched. They can also analyze historical information to determine, for instance, where car accidents tend to occur. In addition, IBM developed a custom flood forecast system for the city... The project cost Rio about $14 million. If it all works according to plan, it could make Rio a model of data-driven city management.


I have a few cursory observations. Apart from the prohibitive cost, many of these technologies, especially the integration of different systems and the development of data analytics that can serve as effective decision-support, are at the initial stages of its evolution. It is therefore to be expected that it will be sometime before the IBM's Intelligent Operations Center catches on and gets scaled up elsewhere.

However, there are several low-hanging fruits in this eco-system. They are low-hanging not only because they can be implemented with limited investments but also would significantly enhance the effectiveness of urban administration and/or improve the quality of life for citizens.

For example, intelligent traffic management systems, which integrate the feeds from all existing hardware - cameras, signal lights, GPS devices in various vehicles, wireless and other police communication systems etc - can be a powerful force multiplier in traffic management. Mobile phone communication signals can be used to map real-time traffic intensity on various city roads (including short-term traffic trends) and the same can be rendered on mobile phone apps to enable commuters to plan their travel more smartly. Similar applications can crowdsource information about utilities related complaints; achieve energy efficiency in all types of lighting and water and sewerage utility motor pumps; cognitively striking data visualization can enhance decision-support systems available for municipal officials, and so on.

Like any new technology, the breakthroughs will come once citizens and urban administrators realize the effectiveness of these systems and the large impact they make on their lives and activities. Given the high cost and nascent technology options, a few quick-wins are necessary to break open this market.

See this and this from IBM and Cisco respectively.

1 comment:

Anonymous said...

Two comments:
Cities are extremely complex "systems of systems" whose management is more in the realm of "wicked problems" as first defined and discussed by Rittel and others in 1970s. They may or may not lend themselves to coordination and control by "rational" systems, and this says nothing of the wildcard in the mix which is the citizens of the cities themselves. The systems are one thing; the users of the systems represent an entirely different level of complexity. Perhaps, cities are too complex and too random to ever really be optimized. See the seminar essay by Christopher Alexander in the 1960s, "A City is Not a Tree."
Two: assume that a city can be rationalized and optimized. Systems are made as efficient as possible. Induced behavior suggest that given the additional capacity --i.e. smoother flowing traffic, etc.-- people will simply take advantage of this capacity by driving more and further. How does this advance the cause of resource conservation, or even make people's lives substantially better?