In Richmond, Va., police deployed
officers in areas with high past New Year’s Eve crime rates, cutting
gunfire incidents by 49 percent and increasing weapons seized by 246
percent. This was accomplished with a third of the officers
typically used and led to a savings of $15,000 in overtime for that one
night.
In Houston, Texas, an analyst monitoring 9-1-1 calls
spotted information about a car on the scene of a Family Dollar Store
robbery, one of a recent spate of such robberies. Using partial plates
and a brief description, he found the car had previously been charged
with moving violations and that the driver wasn’t the registered owner
and didn’t live where the vehicle was registered. The analyst found the
driver’s address and got it to the officer, who went there immediately
and arrested the suspect—solving the series of robberies in under an
hour.
In Erlanger, Ky., an officer in the field quickly solved
an indecent exposure call from his car by searching past records. On
another occasion, when a burglary occurred in a neighboring community
close to the city boundary, supervisors immediately knew and sent units
to patrol the area, potentially preventing further thefts—without having
to add personnel and raise taxes.
These are just some of the
stories police in these cities tell about what they are accomplishing
with business intelligence technology that provides decision-makers with
a 360-degree view of crime and enables them to solve and prevent crime
quickly and effectively while containing costs, including reducing
overtime.
Predictive Analytics Seasoned
police officers know that weather affects crime. In Richmond, Va.,
though, even the greenest officer can know exactly how it affects the
city’s crime patterns and can respond accordingly. In 2004, Richmond,
population 220,000, was ranked the fifth most dangerous city in the
United States. The police department implemented a new information
system for predicting crime and reduced its violent crime rate by double
digits two years in a row. In its first year, the system significantly
reduced incident rates of murder (32 percent), rape (20 percent),
robbery (3 percent), aggravated assault (18 percent), burglary (18
percent) and auto theft (13 percent). All major crime rates have
continued to drop consistently, with homicides dropping by 42 percent
and commercial robberies dropping by 45 percent in 2008.
The
system, which won Gartner’s 2007 BI Excellence Award, provides a
sophisticated data model of criminal activity using elements defined by
the Richmond police to predict future criminal behavior. The entire
department has easy access to predictive crime analysis, data-mining to
correlate past and present data, reporting, and GIS capabilities to view
specific types of crime for a given area and perform crime mapping and
analysis.
Richmond PD worked with Information Builders for integration, analytics and reporting; with
ESRI® for dynamic mapping display; and with
Pictometry International Corp.
for detailed pictures and dynamic geographical displays of locations
and surrounding neighborhoods for reported incidents. They also worked
with analytical software vendor
SPSS Inc., an
IBM company, for data-mining capabilities.
Crime
analysts can now correlate present and past data back to five years,
analyzing arrest records, motive, and type of crime at a particular
location based on the day, time, weather, moon phases, pay days,
holidays and city events. Analysts found, for instance, that certain
property crimes were better indicators of likely sexual assaults than
the presence of convicted sex offenders.
Richmond PD uses this
insight to optimize police resources to deter crime. Officers receive
the most up-to-date information available, along with a screen of
predictions of crime hot spots they can access before a shift. For
example, a high rate of robberies occurred on paydays in Hispanic
neighborhoods, where fewer people use banks and where customers leaving
check-cashing stores were easy targets. Elsewhere, clusters of
random-gunfire incidents occurred at certain times of night. To protect
these neighborhoods, extra police were deployed when and where incidents
were predicted.
Real-Time Crime Center Houston,
the fourth largest city in the United States, has a population of 2.2
million with 5,200 police officers patrolling its 700 square miles. With
information as dispersed as its population, the Houston Police
Department (HPD) needed a single version of the truth and fast, secure,
simple access to accurate, useful data. It met the challenge with a
real-time crime center built in less than three months using
Information Builder’s
business intelligence and enterprise integration solutions for report
writing and data access via a dashboard and ESRI for GIS mapping.
Awarded
a Vision Award for Business Impact in the category of “Advanced
Business Intelligence” by BeyeNETWORK, HPD’s crime center provides
real-time data analysis on more than 15 years’ worth of information and
quickly disseminates it to the field, focusing the whole organization on
its core mission of fighting crime. The system is improving police
procedures and officer and community safety, reducing the time to solve
crimes, enabling better deployment of officers to hot spots, and using
overtime more efficiently.
HPD tied analytics associated with
crimes to incident locations, providing insights into specific incidents
and trends. For example, those dealing with a crime scene can analyze
all the statistics around the area, looking at what has happened within a
time frame and even seeing parolees and sex offenders in the area.
Users can also zoom in on a neighborhood, block or street, and see
houses, cars, yards and fields.
In addition, crime center
analysts have real-time access to Houston 9-1-1 calls, where they watch
on a rolling screen as calls and messages go between the dispatcher and
the car. Analysts work 24/7 analyzing data, generating reports on the
fly, and providing information to officers in the field. Recently, for
example, the emergency center received a disturbance call, and while the
officers were being dispatched to the scene, analysts ran the name,
found out the man creating the disturbance was a fugitive, and alerted
the officers, who arrived at the scene knowing he was a fugitive and
arrested him.
Every manager has access to KPIs and performs
analysis. At weekly meetings, the chief, for instance, can ask how many
armed robberies happened compared to the previous month in a specific
division and access statistics immediately. These meetings are
PowerPoint-free, with commanders productively collaborating on
strategies and immediately accessing all relevant data.
Integrated, Real-Time Search As
part of the Greater Cincinnati-Northern Kentucky metropolitan region,
the Erlanger, Ky., Police Department coordinates with police and fire
departments in more than a dozen communities. Recently, the community
experienced a 5–10 percent annual increase in calls, stretching police
resources to the limit.
To better serve its 17,000 residents
with a 45-member force, Erlanger PD created a new real-time information
system with integrated search capabilities. The new system enables law
enforcement personnel in 10 adjacent police agencies throughout Northern
Kentucky, including 150 patrol officers, to share up-to-date
information from 19 government agencies. The system combines current
crime data from the agencies, linking formerly unrelated information
about suspects, incidents, arrests and crimes. It also merges current
data with crime records and incident reports stretching back more than
five years.
Accessed through a simple Web-based portal
interface, the new system provides police with real-time views of
incidents, arrests, 9-1-1 calls and other events throughout the dispatch
area. Erlanger PD worked with Information Builders, using its WebFOCUS
software and ESRI’s geographic information system (GIS) software to
create this interactive, real-time crime portal in less than three
months.
All cities and agencies connected to the system input
their respective Records Management Systems (RMS). Officers can also
enter notes into the system directly from the field. Integration
technology updates the search index every 15 minutes with crime records
from Erlanger’s computer-aided dispatch (CAD) system and RMS. Users
search both dynamic BI content and structured and unstructured data
sources through the universally known Google search paradigm. By using
simple keyword searches, they can locate key facts and then follow links
to execute reports and access information in the format of their
choice, including Excel, PDF, HTML or XML. With this capability, even if
they don’t know exactly what they are looking for or where it is
stored, officers can find what they need via a Web browser.
Now
it is much easier to detect patterns quickly. In the past, for example, a
string of burglaries involving iron and steel became widespread
throughout the county without individual police departments recognizing
the pattern until some stolen goods showed up at a recycling center.
Today, as soon as an officer enters a keyword such as “metal” or “iron”
into the search index, all the related incidents would be linked in a
one-page report, making the pattern readily apparent.
The system
makes getting data practically instantaneous and delivers information
in appropriate ways for each audience, generating real-time search
results for police officers in the field and delivering KPIs for
supervisors at headquarters. Patrol officers access the system via a
browser-based application on the Mobile Data Computer (MDC) in their
vehicles. For example, if an officer stops a speeding car and performs a
quick license plate search, the system could display a police report
from earlier in the day in a neighboring city involving a hit-and-run
incident, even if the witness at the scene only got part of the license
plate number. A suspect in two crimes could be apprehended through a
small piece of shared data.
Dispatchers and supervisors can view
a different slice of crime data through dashboards at headquarters that
display and drill down into KPIs, including crime activity by city,
current crime alerts, and a summary of arrests and incidents sorted by
categories (such as arson, assault, burglary and criminal mischief). The
new crime portal gives supervisors much more power to prevent or
quickly solve crimes by efficiently deploying the police force based on
patterns that become apparent when records are accessible and grouped in
meaningful ways.
Proactive, Predictive, Intelligence-Led Law Enforcement The
Richmond, Houston and Erlanger Police Departments faced a situation
common to many police departments: data from different systems—and
sometimes other law enforcement agencies—not talking to one another and
being difficult and time-consuming—if not impossible—to access. All
three departments built customized solutions using off-the-shelf
components in a short time to put accurate, timely, targeted information
in the hands of decision-makers—from the police chief to dispatchers to
officers in the field. Most importantly, by making their processes more
efficient and effective with intelligence-led policing, all three saw
immediate and continuing results that help them make their communities
better places to live. ¦ Kevin Mergruen is Information Builders’ vice
president of corporate sales for business intelligence products.
