For small cities, taking steps toward predicting crime

It's not just big cities that are interested in predicting crime -- it's the little ones, too. Often, they're the ones leading the charge. We talk to IBM's director of public safety, Mark Cleverley.
Written by Andrew Nusca, Contributor

Can you predict crime?

Police officers are beginning to think so, thanks to new, robust technology available to cities.

But it's not just the biggest cities that are interested -- the little ones are, too. Often, they're the vanguard of high-tech public safety, even if they don't realize it.

We spoke with IBM's director of public safety, Mark Cleverley, to learn more about his company's recent work with U.S. law enforcement.

SP: How did IBM come to begin working with law enforcement?

MC: The foundation for the "Smarter Planet" initiative, and beyond that Smarter Cities, was that we made three fundamental observations about technology and society and the way people behave.

We call it the "Three I's"

Instrumentation: the increasing number of devices able to communicate, like RFID tags and [traffic] cameras and smartphones and sensors that can tell us about the physical world.

Interconnection: it's getting a lot easier to shift that information around from place to place to system to system to object to person, and standards.

Intelligence: the very significant analytical capability that we can use to ingest this volume of data and analyze it rapidly in time to make a difference.

If you think about the rise of consumer electronics, almost overnight becoming instruments of commerce, that's paralleled by the rise of social networking -- how people interact with each other, including criminals. And that [can require] access to tremendous computing power, even though they don't realize they're accessing it. Societal changes occur in parallel to these observations.

When you boil it down to public safety, people are trying to use technology to solve big societal problems. Public safety is a very important piece of the bigger puzzle. For medium and small cities, it's a very big proportion of their spend. It's also somewhat connected to the economic cost of that city; a safe city is easier [in which] to generate more money.

So we started to pull together the threads within IBM that are working on law enforcement and emergency response management.

Where people invest in law enforcement demonstrates the major initiatives they're trying to improve. The number one issue is fragmentation of information, particularly in the U.S. but seen around the world. There are so many agencies that have evolved with different cultures; a lack of sharing impacts that. Sometimes there are process barriers. Sometimes there are even legislative barriers.

The second point is building that information into a trusted situation. Is the information complete and accurate? For police, that might be the identity [of an individual]. We've had situations working with identity insight software, designed to remove these ambiguities, where there are apparently 500,000 individuals -- yet in real life, there are only 300,000 specific individuals.

The third piece is around mobility. The ability to shift this information in usable form to any device anywhere in a manageable way. The connectivity and management of that is important. Also, we think information design is important -- what information should you be sending to a police officer at the beginning of a patrol? Handheld devices have been given to officers, but they haven't really been adopted. The information design has been lacking. It would be more useful, for example, to get information on five guys I'm likely to run into that day, based on recent events. Increasingly, police officers bring their own devices.

The fourth piece is that these agencies have always spent resources on reacting more quickly. We're working on moving toward being able to predict what is likely to happen, in terms of crime, in terms of natural disasters. An extra day ahead of a snowstorm helps you.

The last piece of it is moving toward a much more unified situation with awareness across all the different agencies that need to be involved, all the way through to recovery, in both the public and private sectors.

Every police officer I've ever met would love to have the real-time crime capabilities that New York has, but they don't have the resources. So we're looking at how to bring them that -- perhaps through a private cloud, perhaps through shared capabilities.

SP: What are the pain points of these departments?

MC: At the top level, there are two major things going on. One is associated more with the so-called developed world: the U.S., Western Europe, Canada -- these countries are under severe financial pressure, at the macro level, austerity and cost-cutting. They have very few extraneous resources; they are under pressure.

If you think about the U.K. police forces, they are under orders by the government to cut their budgets 25 percent in four years. How do you cut and still maintain the level of service you've committed to? That's a big issue.

If you look at the developing world, the so-called "growth countries," I see a different dynamic there. They're being driven by the need or desire for rapid modernization. To leapfrog to where the leading practices in New York, Chicago or Los Angeles are. They're taking on fairly significant new projects. A classic example is the city of Rio de Janeiro, their emergency management center came about because the mayor was concerned about the landslides that were happening a few years ago. A lot of people died in April 2010. So we brought together [urban] planners and linked them to the hospital systems.

Across the board, governments generally share information between them about what they're up to. Sometimes that results in people seeing a particular operation and saying, "We need that." An example is the New York City real-time crime center, which evolved over years. That center is constantly hosting visits from around the world. Police chiefs look at that practice and see it as a world-leading example.

It's becoming apparent that there are many more areas of crossover. I'm talking to a team in Nairobi around traffic congestion, which turns out is linked to public safety because there are dangers associated with that traffic.

SP: How are smaller cities approaching this technology?

MC: In the U.S., even the smallest police agencies now have online systems for computer-aided dispatch, records management, and so on. But they're still very fragmented.

In some countries, it's still very paper-based. You'd be surprised -- cities in India, China and Russia still do a lot of their basic work on paper forms. The kind of solution you need to bring to bear in those situations might be more primitive -- case management, document management.

I remember a conversation early on where I was at a conference where I met the mayor of Calcutta. He sat patiently through the presentations and then said, "Well, my city has five million people living in cardboard boxes." What that indicated to me is that we're looking at these societal problems, but we're assuming that these places are more developed than they actually are. It does put into perspective the sort of resources that these places have.

It's being able to take large volumes of text records and being able to automatically generate linkages: a weapon, a vehicle, a location. Can we build a model of that jurisdiction which actually predicts what sorts of crime are likely to happen? Weather, time of day, is it pay day, is there a big rival high school football game? Predictive analytics is one of the more promising things here.

SP: And for the mayor of Calcutta? How does IBM work with a city that clearly can't afford its products?

MC: We recognize that all not interventions in cities have to be sales or technology. One sign of that is the philanthropic aspect -- the Smarter Cities challenge grant. The grants are made not of money, but people. We ask them to write applications for grants describing a particular problem that they want to address. It might result in recommendations that have absolutely nothing to do with technology; it often is about getting different departments to think differently.

The second thing is, we're pretty intense about thinking about how we deliver solutions -- bringing the best to the rest. Large but resource-poor cities, or smaller cities. Are there ways to bring private cloud technology to deliver shared video technology to deliver analytics? Once you work through issues of where data can and can't reside...

The other place we're engaged with now is collaborative efforts. Consortia and public-private partnerships to help cities move forward. Funding from multilaterals, like the World Bank.

SP: What are small cities asking for?

MC: If you take a relatively small city like Memphis -- population 646,000 -- they were a very early adopter of predictive analytics and crime modeling, and they've had extremely notable success in reducing crime. That's one example of a medium-sized city making leading edge progress.

In the early days, Richmond, Virginia and Edmonton, Canada had similar approaches.

The medium-sized cities are often where innovation rises because they don't have the bureaucracy and overhead of a very large city. In New York, take congestion pricing: it wasn't a bad idea, but nobody could build the right constituency to make it happen.

Memphis, St. Louis, Baltimore, Houston -- those kinds of cities have the problems that could really benefit from the next wave of investment.

One of the reasons we embarked upon this initiative is because we weren't good at engaging with city governments -- but paradoxically, that's where national problems come closest to the surface. We've got to solve them at the city level.

SP: Are cities a focus just because they're in vogue right now? What about regions?

MC: You do see a lot of regional initiatives. What is Beijing: a city, or a region? In some cases, we do see regional projects being established. Germany has a very strong regional organization. Generally speaking, we have to be careful to position everything in context of that country or region. In many places in Europe, people have grown up with a national police force. That's a different kind of engagement than how you'd approach it in the U.S.

The U.S. has grown up in a very particular way with regard to its public safety. There are 20,000 public safety agencies in the country. Compare that to France, where you have a national police force. There, it's almost a military-like structure.

This post was originally published on Smartplanet.com

Editorial standards