• Investment World
• eWorld
• Brand Line
• Mentor
• Life
• Brand Quest
• The New Manager
• BL Club
• Smartbuy
• Books
• Gallery

Stocks

• Quotes
• SE Diary
• Scoreboard
• Open-End Mutual Fund

Cross Currency

• Rates
  
  

Shipping

• Ports
  
  

Archives

• Yesterday
• Datewise
• Resources
• In Focus
• In Depth
• Events 2007

Group Sites

• The Hindu
• The Hindu ePaper
• Business Line
• Business Line ePaper
• Sportstar
• Frontline
• The Hindu eBooks
• The Hindu Images

Raghuvir Srinivasan

Widening the horizon of technology.

If Al Capone were living in New York City today, he would do his utmost to cover the scars on the left side of his face and keep his nickname “Scarface” top secret. For, both will show up on the massive computerised database maintained by the New York Police Department (NYPD) and give him away.

Indeed, it may not pay anymore for even small-time criminals in the Big Apple to flaunt their tattoos proudly, thanks to IBM and NYPD's efforts to develop predictive analytics using the massive amounts of data that the latter generates everyday. The largest police department in the world has a real-time crime database about people who have committed all kinds of crimes — starting from silly stuff such as writing graffiti on the walls or littering the road, to more serious offences such as theft and murder.

“Every time they gave a ticket, they got data and the reality is that people who do little crimes do commit big crimes,” says Frank Kern, Senior Vice-President, Global Business Services, IBM. Big Blue is working with the NYPD on the crime detection centre's database which will warehouse massive amounts of data and eventually move into predictive mode.

A “tattoo database” that will have details of the tattoos sported by criminals; a nickname database and even a database of “connections”, where you can identify a person's friends or relatives are some examples of the types of data warehoused by IBM and NYPD. If this is not enough, New Yorkers have been asked to send video clips or photos of crimes witnessed by them to the ‘911' service of NYPD.

If warehousing such data is a challenge, what is even more so is using it to predict criminal activity. IBM's consultants, programmers and researchers are working in tandem with NYPD to take the analytics to the next level of prediction.

“They (NYPD) asked us to work on video recognition. They have put in cameras in Lower Manhattan and now they will be getting massive video feeds with which you can do facial recognition or portions of the face such as ears, eyes…. We can look for patterns in a crowd to spot trouble or terrorists. Our research people are working on this,” says Kern.

Firefighting, with technology

If IBM's work with NYPD sounds exciting, the job that it is doing for the Fire Department of New York (FDNY) may appear less so but it is certainly no less significant. Late last year, Big Blue bagged a four-year $25 million contract with FDNY to build a risk-based inspection system for FDNY.

It will be IBM's job to help the FDNY find a way to use limited resources in a strategic way through the use of predictive technology. The city of New York has 9,00,000 buildings half of which are residential and cannot be inspected owing to privacy laws. But the remaining half is still too many for FDNY's 16,000 strong workforce to inspect on a regular basis.

“With this manpower, the best case will be that we inspect a building once a decade. But to prevent fire through inspections you need to pick out buildings for inspection cyclically and IBM will help us by using the database to identify buildings that are dangerous,” says William B. Eimicke, Deputy Fire Commissioner, FDNY.

IBM and FDNY have together created a risk-based model that will take into account various parameters such as materials used in a building, existence of fire alarms, the economic and demographic character of the building's neighbourhood, age of the building, its use, and so on, to identify the riskiest buildings that will need periodic inspection. Even factors such as whether the building has figured in a crime before will be taken into account to assess its risk to fires.

Once identified, the fire-fighting unit responsible for the neighbourhood will be alerted about inspections due on a daily basis through a computerised system. The loop will be complete with Eimicke fed back data on whether the responsible unit has done the inspection on the given date.

“This analytics project will identify susceptible buildings and fix them in advance. So we're changing fire-fighting from putting out a fire to preventing one. It sounds simple but it's a fairly dramatic change,” says Kern.

Come December, the project will enter the pilot stage with six fire-fighting units in New York being covered with city-wide rollout planned for Spring 2010.

But will this not have an impact in terms of jobs as the number of firemen required will fall?

Eimicke concedes that but says the focus will shift to medical emergencies which, he says, is a “growth area” because of the ageing population. He has the last word: “Fire vehicles and ambulances are sent simultaneously in an emergency and we've found that fire vehicles reach at least two-three minutes before the ambulance does!” Now that's an indirect benefit of IBM's analytics for those in medical emergencies in New York.

rags@thehindu.co.in

More Stories on : Software | Security | Technology

Article E-Mail :: Comment :: Syndication :: Printer Friendly Page

Copyright © 2009, The Hindu Business Line. Republication or redissemination of the contents of this screen are expressly prohibited without the written consent of The Hindu Business Line