Identifying Needs

When the FBI decided to upgrade its ID system, the bureau surveyed the system's disparate users to help it set requirements.

Eric Butterman

The first step in solving a problem is to identify it. But for the FBI, identification actually was the problem.

When the bureau began plans more than four years ago to upgrade its ID systems and create the Next Generation Identification system (NGI), it faced a challenge: How could it get the most compelling information about what changes the system needed?

The Integrated Automated Fingerprint Identification System (IAFIS) has been the central ID tool for the nation’s law enforcement organizations at all levels of government. The bureau issued the solicitation for the 10-year, $220 million NGI project this month. But to get to that point, it spent nearly two years developing the requirements for the new system and relied on an extensive user survey to define and refine those requirements.

Although fingerprinting is an effective means of identifying suspects, it can be hard to share the information. Facial IDs can be effective, yet not every law enforcement agency has this ability. The FBI realized that if it was going to develop solutions and new applications for NGI that would work for all levels, it would have to get input from users at all levels.

Enter the NGI 1,000-user canvassing initiative, which involved extensive surveying, data collection and committee meetings — and, more important, led to results, says Kim Del Greco, section chief for the FBI’s Criminal Justice Information Services Division.

The biggest challenge was the timeframe, Del Greco says. “When you begin a development project, you want to get to an acquisition phase, so we had to build something new to canvass requirements that worked on many levels, including the system level.”

Step 1: Create a Team

To begin the requirements planning process, the FBI created an advisory board of participants from across the IAFIS user spectrum.

The FBI recognized what it didn’t know and used “a whole shared management concept,” says Mike Lesko, deputy administrator of crime records services for the Texas Department of Public Safety. “I have had dealings with other agencies, and the FBI embraced our help. ”

To help look at particular issues, the FBI created subcommittees. Lesko led one that tackled state and local IAFIS use. “I gathered a number of local and state entities to see how records go up to the FBI and come back for fingerprint submissions. We suggested cases and had oversight of biometric specifications,” he says.

Step 2: Figure Out Whom to Survey

To do this, the NGI team realized stakeholder input would be critical. “We scaled out and mapped out how many users were coming from the state level and how many we’d want to survey from the bigger states because they’ll obviously have more need,” says Del Greco. “Gathering information was done also with the idea that federal users have additional legislation to apply.”

Step 3: Collect Responses

The team’s survey asked respondents how they use IAFIS and also to identify what other identification tools and information they would like to have from a new system.

The FBI proactively canvassed approximately 111 state and U.S. territory agencies, such as state identification bureaus and crime labs, 54 federal entities, 18 authorized non-criminal-justice agencies, and 10 special interest groups that included organizations such as the National Institute of Standards and Technology and the SEARCH Group (www.search.org).

Step 4: Break Down the Results

Using an online application for electronic capture and analysis, the NGI team did multiple cuts of the responses. The bureau decomposed the stakeholder responses into detailed user requirements around functional and nonfunctional requests. The team then broke out the results into specific system requirements.

The FBI established the categorization criteria, and then the advisory board approved them and ranked the relative importance of the criteria. The bureau developed its survey and data analysis approach based on industry standard best practices, Del Greco says. Once it had created multiple reports around the category criteria, the acquisition team used them to draft its request for proposals.

Lessons Learned

Lesko says that in the end there wasn’t much to improve upon in the canvassing process, but the project did result in information that all the participants can build upon for other projects.

“Users are stakeholders and have vested interests, and they want to be solicited,” he says. “There was a thorough questionnaire, and it addressed areas for operators at the local level. The FBI also gave latitude for information that wasn’t germane to the questions but was relevant to identification processing. They made us believe we weren’t wasting our time.”

Bill Casey, a deputy superintendent for the Boston Police Department, says the approach helped him see that the users need to improve how they provide feedback to the FBI.

“We have to do a better job of telling the FBI where the rubber hits the road,” he says. “This whole process has made me cognizant of the subtle differences from state to state. In Massachusetts, we have a record of informational law of who you can use criminal justice information on, while Florida can put tons of strong facts about a criminal right on its Web site. We need to have a better idea of which states have less room to maneuver and figure out programs that work for those states that have minimum options.”

The FBI plans to mirror the survey approach for other systems projects. “A project will only be as successful as the user input,” Del Greco says. “You have to make sure the right leaders are involved, and you’ve communicated what should be expected of them.”