In an editorial in the New York Times Thursday, the paper calls the 2007 Secure America Through Verification and Enforcement Act, " a bad idea compounded by the notoriously bad state of federal government records."

The act would, among other things, "force all workers, including citizens, to prove they have a right to earn a living," by relying on the Social Security Administration to verify Social Security numbers for workers, the paper contends. The problem is that one SSA database has a 4 percent error rate, which would mean possibly thousands of workers would face firings and discrimination.

Other federal databases contain errors. The inspector general at the Justice Department reported last year that the Terrorist Watch List, which is used to screen 270 million people a month to identify possible terrorists, has a large error rate. "In an examination of 105 records, for example, the auditors found that 38 percent of the records contained errors or inconsistencies that the [Terrorist Screening Center's] own quality-assurance efforts had not found," according to a Washington Post article.

As the federal government relies more on information technology to support critical decisions, the importance of how clean its data is rises.

How confident are you that your data is error free?

The Risk Factor blogger Bob Charette, a risk management expert who consults with federal agencies on risk management, picked up yesterday's story on the deep trouble that the Census Bureau's handheld computer contract is in. In his blog post, he questions the credibility of the Capability Maturity Model Integration (CMMI®). Harris Corp.'s Government Communications Systems Division, which is the prime contractor on the $600 million handheld contract (now likely much more than $650 million after all the costs from changes, errors and delays are included), has a Maturity Level 3 rating. "The Level 3 rating denotes superior process maturity within the division's program management, engineering, quality assurance, and other disciplines, and achievement of this rating has become a competitive differentiator on many government programs," Charette quotes.

Charette wants to know: "At the very least, I think the division's CMMI rating may need to be re-evaluated, or maybe better, the U.S. government better start looking at what, if anything, SEI CMMI Level 3 actually means in practice."

Or it could mean, the customer, the Census Bureau, put too many demands on Harris -- so many, in fact, that no maturity designation, no matter how high, could have avoided the very problems that now threaten the viability of the U.S. census.

The Census Bureau has had trouble managing the costs, time lines and, most important, the performance of a contract to develop a handheld computer to collect data during the 2010 census. The cost of the contract has increased from its original $600 million to $647 million, according to a General Accounting Office report released today. If all related costs due to the handheld contract's delay and mismanagement are taken into account, GAO estimates the increase in costs for the 2010 census could range between $600 million and $2 billion.

While those overrun costs are high, many government information technology projects (and private-sector IT projects) have suffered similar fates -- with little or no repercussions for the agencies. But Sen. Tom Coburn, R-Okla., and a member of the Committee on Homeland Security and Governmental Affairs, has suggested something new that could set a precedent for other agencies. In a committee hearing held today on the problems with the handheld contract, Coburn suggested that any cost overrun in the contract be covered by the Census Bureau cutting the budgets for programs in future budgets. This is what he said:

For years, the Census Bureau has estimated that the 2010 count will cost between $11.3 billion and $11.8 billion. I hope that the Secretary of Commerce will work to ensure that the cost does not increase beyond that, even with these trying circumstances. However, let me be perfectly clear -- if the costs go over that amount, taxpayers should not have to subsidize this mismanagement more than they already have. If more money is needed, I fully expect that the department and the bureau will work internally and with [the Office of Management and Budget] to find offsets out of programs that already exist.

The Census Bureau's total budget for fiscal 2008 is about $1.5 billion, with larger budgets coming at the end of decades to pay for conducting the decennial census. Using even the conservative estimate of a $600 million cost overrun in the hand held contract would present a financial challenge, to say the least.

The administration’s top IT official bid an early farewell to government and industry IT workers at the 2008 Information Processing Interagency Conference in Orlando, Fla., Monday before announcing the winners of government project management awards.

Acknowledging the approaching end of the Bush administration, Karen Evans, administrator of the office of e-government and information technology at the Office of Management and Budget, called her fifth keynote at IPIC “bittersweet.” She then acknowledged the work of agencies to achieve the goals of the e-government initiatives, which identified several governmentwide programs to integrate agency operations and information technology investments.

“It isn’t work OMB has done; the work is done by vendors that help the agencies and the agency [IT administrators],” Evans said.

Recipients of government project management awards, some of which Evans announced Monday and others that will be announced at a Tuesday session, were recognized for programs that demonstrate excellence in project management:

Cost Savings/Cost Avoidance
Winner: NASA’s Shared Services Center
Winner: The Office of Personnel Management’s Human Resources Integration

Retooling the Infrastructure
Winners: The Energy Department’s Pacific Northwest National Laboratory Campus Camera and Emergency Call Station System; and the Interior Department’s 104 Mainframe Efficiency Improvement Project

Service-Oriented Architecture
Winner: DHS’ U.S. Citizenship and Immigration Services Enterprise Service Bus

Digital Trust-Infrastructure Security
Winner: GSA’s Managed Service Offering USACCESS Program

Identity Management-Biometrics
Winner: Federal Bureau of Investigation’s Quick Capture Platform

Delivering Mission Services/Practical Innovations
Winner: Bureau of Alcohol, Tobacco and Firearms’ ATF Knowledge On Line

Delivering Mission Services/Practical Innovations
Winner: Small Business Administration’s Business Gateway Initiative

The following item was posted on the Blog "The Agile Mind," written by Anne Laurent, who gave permission to have the item posted in its entirety in Tech Insider.

The military services have been early and avid government adopters of gaming technology and especially software platforms. The Army has had tremendous success with its recruiting game, America's Army, and follow-on training modules built on the same game engine. The Defense Advanced Research Projects Agency and Sandia Laboratories have created wildly successful trainers for languages -- chiefly Iraqi Arabic -- culture and non-kinetic (civil-military) operations.

I've written about DARPA's projects and their Godfather, Ralph Chatham, for Government Executive magazine. I caught up recently with Chatham, who just left DARPA at year's end. Among the many strands in our rich and fascinating discussion, we talked about a very exciting DARPA project that could revolutionize the way the military -- and the rest of government -- uses games, as well as wreaking wide-ranging effects on the way games are created in the commercial world.

In March, Chatham expects to see a first version of DARPA RealWorld become available for use in the field. What's the big deal? Well, the huge speed bump to military use of games for training is that service members haven't been able to easily alter them to accurately represent the terrain, buildings--outside and inside--and vehicles they confront in the field. Real World is designed to be truly user authored by not-so-technically-adept soldiers, airmen, sailors and Marines on the ground. That means Real World can become a real mission rehearsal tool.

Daniel Kaufman, the RealWorld program manager, says his goal is to be able to build simulations without programmers. This "dictates a new approach to getting software requirements," he told the audience at DARPA Tech 2007, the systems and technology symposium held in August in Anaheim, Calif. "The 20 meetings to write the 100-page RFP to generate the 1,000-page specification to find a product that will not be delivered for four years has consistently failed," he continued. So he set out to build tools and capabilities so warfighters can create applications when they need them. Take a 19-year-old soldier in the field, Kaufman said:

He’s out on patrol in a rocky canyon in Afghanistan and some OPFOR pops up and shoots at him because that’s what an opposing force does. Our warfighter engages, the OPFOR vanishes, and our Soldier returns to base to be debriefed by his commanding officer. Our soldier gets out his laptop – and, voila! On the screen appears a scene that is an exact 3-D recreation of precisely where he was in that canyon. Not generic terrain – this is exactly his patrol and exactly his location.

Within seconds, our soldier is dragging-and-dropping:

“This is where I was; this is where my buddy was; this is where George was; this is where the HMMWV was, this is where the sniper was, and this is where we got shot, sir.”

Notice that I said he does it. There’s no software guy; there’s no writing down specs. He does it, and within seconds it’s right on his laptop screen and it’s exactly correct.

If you think about it, in that one small instance, four very important things have taken place: RealWorld has become an after-action review tool, a mission planning and briefing tool, a mission rehearsal tool and a training tool.

Imagine recording this whole sequence, and then sending it back by e-mail to Ft. Polk and Ft. Lewis, and Twentynine Palms, or anywhere else. And instead of trying to tell a kid back at a U.S. training base, "Look here’s 100 pages of doctrine that explains how you are supposed to handle an IED, and here’s a PowerPoint slide, and here’s a satellite map, and here’s a contour map," we put him right there!

Now training takes on a whole new meaning. Our stateside soldier is not working with, "Here’s a square: imagine that’s you, and imagine the bad guy is this circle over here." We’re saying, “In 90 days, you’re going to be there. Work with this simulation and figure out what you would do. Because if we have not gotten that sniper – who really does exist -- in three months, odds are he will still be out there and it will be your job to go get him.”

OK, so that's a revolution in military simulation, but what about overturning commercial game creation? Kaufman's prime contractor, Total Immersion, is making a bet by developing RealWorld for very little money. The company is getting its R&D paid for and gets to hang onto the real-time mission-rehearsal building tools it is creating. Since it now costs $20 million to $40 million to build a computer game, companies only invest in those that appear to have "blockbuster" written all over them. But what if a company developed a set of tools it could both use and license relatively inexpensively to others to use to create games quickly? Kind of blows open the whole game economy, eh?

More on all this to come, but for now, it's worth noting that before DARPA, Kaufman worked for DreamWorks Interactive, a joint venture between Microsoft and DreamWorks SKG, where he was involved in creating games including Goosebumps, The Neverhood, Jurassic Park and the precursor of what was to become Medal of Honor. Before that, he was an attorney with Brobeck, Phleger & Harrison (Palo Alto, Calif.), where he had the largest game company representation in the United States, handling the EA/ABC joint venture, Spectrum Holobyte's management buy-out and merger with Microprose, which led to an IPO, the formation of Crystal Dynamics, and the formation and subsequent sale of Humongous Entertainment for $76 million. Oh, and the CIA's venture catalyst, In-Q-Tel, once commissioned him to look into how gaming could help the CIA train, too.

Smart development, smart acquisition, smart partnering with the private sector and smart risk taking. Watch out big, entrenched military simulation companies!

Problems and skepticism surrounding the Army's $200 billion Future Combat System (FCS) have been discussed for years now. (Government Executive Magazine ran its latest in-depth assessment last year, and The Washington Post ran a critical article this week.) In the Post article, reporter Alec Klein, quotes an executive from FCS prime contractor Boeing as saying the original estimate of the number of lines of code has always been 55 million -- not the 33.7 million lines of code that the Army estimated in 2003. (That number has now increased to 63.8 million.) Boeing FCS program manager Dennis "Muilenburg ... said that the original software estimate was 55 million lines of code, not 33.7 million," according to the Post.

That statement caught the eye of Robert Charette, a risk management expert who consults with governments, including the Defense Department, and companies worldwide. In his blog "The Risk Factor," he wrote in a recent post:

The reason I find this [Boeing's assertion that 55 million lines of code was the original estimate] curious is that the 33.7 million lines of code estimate has been around for several years, and appears in congressional testimony many, many times. That number gave lots of folks pause in 2003, since the Army claimed at the time that it would complete FCS in five and a half-years. Questions were raised then about whether that amount of code could be developed in that time frame, but the ever-confident Army said it could be accomplished.

I have never heard or seen that 55 million lines of code number ever mentioned before this article. If that was the true estimate at proposal time, did the contractor and the Army "forget" to let Congress, the Governmental Accountability Office (GAO), and a whole bunch of other people know the true system size so that they wouldn't ask questions in 2003 like, "Tell me again how you plan to develop and integrate an average of 10 million lines of native and commercial-off-the-shelf software per year over each of the next five years?" "Can you point to any military software-intensive development of 10 million lines of code successfully completed in 5 years?" "Can you prove you are not legally insane?"

The Hill has become so concerned that the Census Bureau's development of its handheld computers the bureau plans to use in the 2010 census is in such trouble that House and Senate staffers have called the bureau's top leaders to a special meeting on the Hill today.

Newly confirmed Census Bureau Director Stephen Murdock and Deputy Director Jay Waite will meet with staffers from the Senate Committee on Homeland Security and Government Affairs and from the House Committee on Oversight and Government Reform. In 2001, Waite created a plan to use handheld computers to collect data from Americans who fail to send in their paper census forms.

The Hill's concern over the performance of the handled computers, which the bureau is developing under the Field Data Collection Automation program, was heightened after Government Executive posted an article about a Nov. 29, 2007, meeting between Waite and Mitre Corp. During the meeting, Mitre presented its analysis of the contract's progress. According to a talking points document obtained by Government Executive, Mitre characterized the contract as being in "serious trouble" and advised the bureau to immediately develop a contingency plan for using paper forms to collect census data in case the handhelds did not work as planned.

Reps. Tom Davis, R-Va., and Mike Turner, R-Ohio, have asked the Democratic leadership on the House government oversight committee to schedule a hearing to investigate the Census Bureau's management of a project to develop handheld computers it plans to use for the 2010 census.

In a Jan. 7 letter, Davis and Turner cite a Government Executive article published Jan. 2 that was based on a Mitre Corp. document characterizing the handheld computer contract as in "serious trouble" because of poor management. The representatives have asked Rep. Henry Waxman, D-Calif., chairman of the Oversight and Government Reform Committee, and Rep. William Lacy Clay, D-Mo., chairman of the Information Policy, the Census and National Archives Subcommittee, to schedule a hearing on the subject.

In particular, Davis and Turner want to know why Census Bureau Director C. Louis Kincannon did not mention during a subcommittee hearing held Dec. 11 the meeting that Census managers had with Mitre to discuss the contract's problems. The meeting with Mitre, in which Mitre delivered an unusually harsh analysis of how Census was managing the handheld contract, as outlined in Mitre’s talking points document, was held Nov. 29, 12 days before the hearing.

The Census Bureau's senior leadership has maintained – rather stridently at times – that the handhelds will work as planned for the 2010 census and that the only problems they have experienced are those consistent with any IT project of this magnitude. They have insisted they do not need to develop a backup plan in case the handhelds do not work. Forming a backup plan to use paper would be too costly, they argue. But the Government Accountability Office (and now Mitre), as well as project management experts, have all strongly questioned the Census Bureau's management of the contract, especially the inability to manage the inherent risks, as outlined in a Government Executive article last summer.

If the handhelds did not work as planned and the Census Bureau had to revert to using paper forms to collect census data, the cost of the 2010 census would increase by the billions of dollars. Already the cost of the census is tracking to be more than double the $6.5 billion cost of the 2000 census. Nevertheless, Congress would spend whatever it had to to conduct the census, because, as one Hill staffer reminds us, the census is a Constitutional requirement.

Update: In his blog, The Risk Factor, risk management expert Robert Charette, who is quoted in the Government Executive article on the Mitre analysis, discusses just what the Census Bureau means when they understand the handheld contract has "challenges."

It's official: The Senate confirmed four new leaders at the Homeland Security Department last night, one of which could play a key role in cybersecurity efforts.

Robert Jamison was appointed under secretary for the National Protection & Programs Directorate. The office is charged with minimizing the department's risk through an integrated approach of physical and virtual threats. Previously, Jamison served as deputy administrator of the Federal Transit Administration, leading a transit security program and Lower Manhattan transportation recovery operation, which was established after 9/11.

Other confirmations included Julie Myers as assistant secretary of the U.S. Immigration and Customs Enforcement (ICE), Jeffrey Runge as chief medical officer and assistant secretary for the Office of Health Affairs, and Ross Ashley as assistant administrator of the Federal Emergency Management Agency.

DHS Secretary Michael Chertoff released a statement on the confirmations this morning.

We think you, the technology manager in the federal government and industry, have a pretty good insight into just what are the hot issues and events that will unfold in 2008 for the federal IT market. Over the past few weeks we've invited you to take an online survey to let us know what you think; we just want to take this opportunity to invite you to take the survey again, if you haven’t.

We are conducting the survey in conjunction with our friends at Government Futures, which is also offering readers a chance to place bets on what’s going to happen in the federal IT community using the prediction markets on Government Future's Web site.

If you have taken the survey and placed your bets, thank you. If you haven't, please visit the site and give us your opinions. The questions cover a number of hot areas, including information security, the next-generation Internet and federal information technology spending.

In January, we’ll host a webinar to discuss the results of the survey and present an analysis of the predictions.

In the December issue of Government Executive, we discuss some trends that IT experts told us would be important. Now, we want your opinion. So, please take the survey and join the government futures market to help us figure it out.

The Army’s key, next-generation battlefield communications systems, Warfighter Information Network-Tactical, earned the dubious distinction of ending up on the Defense Department’s list of programs that experienced cost increases of more than 15 percent or schedule delays of six months or more. These programs are reported to Congress.

For the quarter ended Sept. 30, the Defense Selected Acquisition Report (SAR) put the cost of WIN-T at $16.4 billion in constant 2003 dollars, up $4.6 billion from the 2003 baseline. That’s nearly a 40 percent increase, for those keeping count.

WIN-T is the network glue intended to stitch together the myriad platforms in the Army’s Future Combat Systems program. The Government Accountability Office reported in March that the Army has restructured WIN-T system development for five years, delaying a production decision until 2011.

I bet new Army CIO Lt. Gen. Jeffery Sorenson was just delighted to find out WIN-T had made the SAR-list on his second day on the job.

Health and Human Services Secretary Mike Leavitt wrote in his blog that he wants to see Medicare and Medicaid and large federal health care providers make e-prescribing “a mandatory part of medical practice soon.”

Leavitt wrote in his blog that a low rate of adoption by physicians has slowed e-prescribing nationwide. “Most doctors haven’t invested in the necessary technology to do e-prescribing,” he wrote. “The reasons are complex and range from a perceived lack of financial incentives to a reluctance to give up the familiar prescription pad. It is not expensive. This change needs to happen, and from my standpoint, sooner rather than later.”

Leavitt did not define what he meant by soon, and I’ve not heard back from HHS asking about it. But the Centers for Medicare and Medicaid services released Nov. 16 final standards for the Medicare e-prescribing program, which covers million of patients. Ray Sass, an HHS spokesman, said he expected these rules to be adopted in less than a year.

Lee Shapiro, president of Allscripts, an electronic health record and e-prescribing software vendor, said cost should be no impediment to adoption of e-prescribing. Allscripts, along with its partners in the National ePrescribing Software Initiative, have offered to provide free software to any clinician in the country who wants to give up their prescription pads and enter the electronic age.

Shapiro said e-prescribing will help cut billions of dollars a year from the national heath care bill and go a long way to reducing the 7,000 deaths a year caused by adverse drug reactions.

The IT managers over at the Internal Revenue Service got some good news this week from an unexpected source: the Treasury Inspector General for Tax Administration. In a report released this week, the IG concluded that the IRS had successfully deployed the latest update to its new database and application engine, called the Customer Account Data Engine (CADE). Developing a working CADE has been a pretty big deal for the IRS; it’s at the heart of the agency’s multi-billion-dollar modernization effort, which has had its history of troubles. IRS agents eventually will use CADE to store and access the 200 million individual and business tax files Americans submit every year. The hope is that the files will be updated daily instead of weekly (improving service by allowing IRS employees to give more up to date informaiotn to taxpayers) and allow for tax returns to be sent out faster.

This milestone was a big one for the IRS because the new release, called Release 2.2, included the most commonly used tax forms, including returns for Single, Married Filing Jointly, Married Filing Separately, and Head of Household, as well as profit or loss from a business (Schedule C), capital gains and losses (Schedule D), supplemental income and loss (Schedule E), profit or loss from farming (Schedule F), and self-employment tax (Schedule SE).

But the IRS pulled it off, according to the IG. Success meant the IRS accurately posted the returns for those commonly used tax forms. The only drawback, the IG reported, was that the IRS processed 11 million returns using CADE instead of the 33 million returns it had set as a goal. The IRS said it missed its goal because it had to postpone launching the new release by two months so it could complete upgrades. The laucnh had one other hiccup that the IG downplayed. The IG caught caught one incident in which the IRS would have sent out $400,000 in undeserved refunds to taxpayers for overpaid taxes, but the error was caught before the payment was sent.

You just knew there had to be an IT angle to the special inspector general reports on procurement abuses associated with the Iraq War. You were right. There is. Seems that the United States spent $38 million to develop a financial management system for Iraq's government. When it stopped working for a month, no one noticed, according to an Associated Press article. From the article:

"According to U.S. Embassy officials, the Ministry of Finance continues to use its legacy system for overall budget and accounting, 'nobody noticed' when IFMIS was down for a month and no one relies on IFMIS to produce reports," [special inspector general for Iraq reconstruction Stuart W. Bowen Jr.] said.

Other ministries, such as interior and defense, have developed their own financial management information systems, and they are not compatible with the new one and cannot transfer financial data from one system to another.

Sound familiar?

In an item posted today in his blog, “The Risk Factor,” risk management expert Bob Charette calls into question OMB's announcement yesterday that the number of IT projects on its Management Watch List had dropped 61 percent – in seven months. “This is truly amazing,” Charette writes. “Sixty-one percent of government IT projects on the OMB watch list, which indicates whether they are well-positioned to execute, all got better at the same time. One can only conclude that the government has found a new, secret way to manage IT project risk.”

The skepticism doesn’t stop there. In an article posted today on Government Executive’s Web site, government project management expert J. Donaldson Frame says, “When I see miracle improvements occur very quickly, I wonder whether the improvements are genuine or reflect statistical artifacts."

And Ray Bjorkland, chief knowledge officer at federal marketing research firm FedSources, wonders how IT projects get on (and presumably then come off) the Management and High Risk lists in the first place.

For the 212 IT projects that came off the Management Watch List, OMB officials said those “agencies were able to adequately address deficiencies and weaknesses identified in these 212 investments by mitigating planning deficiencies, or in some cases, providing and completing additional documentation supporting their management activities.” No word on how well the projects are meeting budget, deadlines or performance measures, which Bjorkland says are the best indications of success in oversight of technology investments.

And the reason given for more IT programs going on the High Risk List? Again, better reporting from agencies, OMB said.

Interesting, better reporting was the reason OMB gave yesterday for the doubling of the number of reported security breaches exposing personally identifiable information. “An increase in reporting isn't necessarily a bad thing,” said Karen Evans, who holds the Bush administration’s top IT executive position at OMB.

This reason given when on the same day, Microsoft reports that phishing scams had increased more than 150 percent in the first six months of 2007 and the number of malware incidents increased 500 percent. Not to mention the 90 percent increase (over nine months) in the number of cyberattacks directed at electric utilities.

It still hurts my head to try to follow this logic. The message seems to be: It's good to know how bad things are. That could be helpful, if you then used that information to develop a plan to fix the bad things. No word on that, yet.

McQ Inc. said Boeing has selected it to provide a family of unattended sensors for the Homeland Security Department’s electronic border fence project, called the Secure Border Initiative Network (SBInet). McQ has a basic ordering agreement to provide unattended acoustic, magnetic and infrared sensors, according to contract information posted on Boeing’s online SBInet Toolbox contract page.

This summer, the Army Research Lab selected McQ’s unattended ground sensors as one of 10 “greatest inventions” of 2006.

In a related development, the House Homeland Security Committee plans to hold a hearing on the troubled SBInet project Oct. 24. DHS Secretary Michael Chertoff threatened last month to withhold payments on SBInet until Boeing fixes problems on a 28-mile pilot project in Arizona. “I'm not going to buy something with U.S. government money unless I'm satisfied it works in the real world,” Chertoff said last month at a congressional hearing.

It seems that most federal projects with a lot of zeros in their budget run woefully way behind schedule and way over budget. But the Energy Department’s Sandia National Laboratories proves big projects can come in way under budget and way ahead of schedule.

Sandia said it completed its $516 million Microsystems and Engineering Science Applications (MESA) project $40 million under budget and three years ahead of schedule. Sandia describes MESA as “a major capital construction activity that will create the facilities and equipment required to design, prototype, and fabricate qualified microelectronics and microsystem components for nuclear weapons.”

A Sandia spokesman couldn't -- yet -- offer reasons why the project was completed under budget and ahead of schedule. However, he did say that Sandia bought two old chip wafer machines from Intel for $25 each. The machines were valued at $7 million each.

Sandia plans to dedicate the final building of the project, the Weapons Integration Facility, located on the Sandia campus at Kirtland Air Force Base in Albuquerque, N.M., Aug. 23.

The 400,000 square foot MESA complex – the largest project in the history of Sandia, which started operation in 1945 as an offshoot of nearby Los Alamos National Laboratory – also includes the previously opened Microfabrication Facility and the Microsystems Laboratory.


















MESA Microlab

The Weapons Integration Facility includes laser, electrical, visualization and computer labs and office space for 375 scientists and engineers. Sandia said the MESA complex will produce ”hardened” electronic circuits and computer chips that can withstand high levels of radiation to insure the reliability of nuclear weapons and other capabilities under even the most hazardous of conditions.

Pooh-bahs scheduled to show up for the dedication ceremony this Thursday include Thomas D’Agostino, administrator of the National Nuclear Security Administration, and
New Mexico’s senior senator, Pete Domenici.

If Sandia promises marching band – and how can you open anything without a band – I may show up.

What is it with computer systems designed to serve the most vulnerable children?

From The Columbus Dispatch, comes another story of a computer system that was poorly developed and puts at-risk kids at even greater risk. The Ohio Department of Job and Family Services last year turned on a new computer system -- the Statewide Automated Child Welfare Information System -- to better track children in foster homes. But when the state began using the system, the foster families who did not have children in their care at the time the system went online were not placed in the foster family database. When children were placed in those families, case workers could not add the family to the computer system. That makes it more difficult to track children put in the care of those families.

As a result, the state runs the risk of losing track of foster children, according to the article. Counties are still being added to the system, but child welfare advocates have called for the state to stop adding counties until the problem can be fixed. The state says an electronic fix is not expected to be available until January, and it doesn't want to stop adding counties to the system because the system already is far behind schedule and over budget. The system has cost $93 million to develop and has had a history of problems and missed deadlines for the past decade, according to the article.

Numerous state and local jurisdictions have been upgrading child welfare systems -- without much success. New York and Philadelphia have reported similar problems with new computer systems developed to better track cases for state child protective services agencies.

Aviation officials in Los Angeles are pretty steamed at the folks at U.S. Customs and Border Protection.

A computer system used to process international travelers coming into the United States was down for nine hours Saturday, creating a backload of 17,000 travelers looking to enter the United States, according to a Los Angeles Times article. Thousands of travelers were stranded on planes for hours. According to the article, Steve Lott, chief spokesman in North America for the International Air Transport Association, explained the airport’s frustration with U.S. Customs this way: “Although ‘we understand that computer systems are not perfect, the frustration is why customs had no contingency plan.’"

LAX officials may be on to something. In June 2004, the Federal Emergency Management Agency issued the "Federal Preparedness Circular," which was sent to the "heads of federal departments and agencies.” The circular presents guidance on how agencies can set up a Continuity of Operations (COOP) plan. According to the circular (emphasis added):

It is the policy of the United States to have in place a comprehensive and effective program to ensure continuity of essential Federal functions under all circumstances. ... All Federal agencies, regardless of location, shall have in place a viable COOP capability to ensure continued performance of essential functions from alternate operating sites during any emergency or situation that may disrupt normal operations.

It seems as if most agencies didn't follow FEMA's guidance because on May 9 President Bush issued National Security Presidential Directive 51 and Homeland Security Presidential Directive 20. The directives mandate that agencies develop a COOP plan “to ensure that Primary Mission-Essential Functions continue to be performed during a wide range of emergencies, including … technological … emergencies.”

Bush's directive obviously came too late for international travelers coming through LAX Saturday. So maybe now's a good idea for a COOP plan to be at the top of Jayson Ahern’s to-do list at U.S. Customs. It was just last week that Ahern assumed the position of deputy commissioner for U.S. Customs and Border Protection – the No. 2 position at the agency. That's one bad first week on the job.

An increasing number of local governments are getting into the business of providing Wi-Fi service to residents who want to access the Internet throughout a city or county. About 385 cities, communities, and counties in the United States have a wireless networking project, with most intended partially or wholly for residential use, according to a recently released report by Forrester Research. (Requires a subscription.)

But Forrester researcher Sally Cohen questions if the investment is worth the cost. Only 27 percent of all U.S. online households use Wi-Fi, and the majority of these users (76 percent) connect to a private Wi-Fi service in their home, not to a municipal or county network provided in, say, parks, libraries, commercial areas or other hot spots.

To make wireless networks a better investment, Cohen advises local governments to do some homework. This includes determining what percentage of residents want a wireless service, how much they may be willing to pay for certain services available on the network, if other services can piggyback on the network such as remote parking payment systems and traffic control video surveillance, and educating residents abut Wi-Fi to increase usage.

The Forrester report, however, doesn't discuss the controversy of local governments providing what telecommunications companies argue is a business that government has no place competing with the private sector.

Around the turn of the latest century, when information security was beginning to get more attention in corporate and government IT shops, one of the first (and most basic) system development best practices proposed by almost every IT project management consultant was to design information security into a system upfront, not after it was tested. In 2001, the National Institute of Standards and Technology issued guidelines that said as much.

The reasons were simple: It's more expensive to include security after development, and, most important, security is typically not as effective if tacked on.

But as Government Executive's Bob Brewin reports, that's exactly what Boeing and the Homeland Security Department have done with its billion-dollar-plus Secure Border Initiative Network (SBInet) surveillance system. As Brewin reports, the Wi-Fi wireless SBInet pilot project Boeing is now testing in Arizona is vulnerable to cyberattacks. Boeing has issued a request for proposals to secure the wireless network, but as any IT manager would have told you six or more years ago, that's not an advisable management strategy if you want the system to be secure.

Why do IT project developers, even the best of them, still -- after six years -- fail to design information security into systems upfront?

A new report from the IT management consulting firm Forrester Research identifies the 10 most common mistakes in project development and how to avoid them. Forrester researchers point out that many of the mistakes are well known. Nevertheless, project developers keep making them, so it is always helpful to review them, Forrester advises.

The top 10 are:

1. Never committing to project success. When a project threatens a business unit, it has a low chance of succeeding. Users in a unit undergoing technological change are not helpful in identifying requirements and are vague about their work. Developers need to work extra hard in showing why the project is useful.

2. Freezing the schedule and budget before the project is understood. Executives tend to hold project developers to the initial, yet sketchy, budget and schedule timelines. Because developers don't understand much about the project at the beginning of the cycle, estimates are no more than guesses. Some developers fund a team to investigate the project's details to create a more informed budget and schedule.

3. Overscoping the solution. Forrester calls this using "a chainsaw to open an envelope." Sometimes a simple tweak or fix -- not a new system -- to an existing application is all that is needed to satisfy a requirement.

4. Circumventing the application development organization. Sometimes a business unit or agency office will hire a contractor to develop an application without informing the central IT office. Integration with the organization's systems can prove troublesome. The application development team must educate the organization's units that it should be informed to avoid technical problems.

5. Underestimating the complexity of the problem. Business users fail to provide the necessary detail for requirements so that the application development office can judge how complex the functionality of the system must be. To avoid this problem, the Program Management Office must deliver applications in stages to receive feedback on whether the system is meeting requirements. That way adjustments can be made earlier in the process, saving money and time.

6. Being stingy with subject-matter experts. When the line of business does not provide enough subject-matter experts -- those who intimately know the business and what a system needs to serve the business requirements -- projects go off track. Without a subject matter expert, application developers will tend to make their own decisions as to what needs to be done to keep the project on schedule. To make it work, a subject matter expert or experts must be freed up from their daily jobs to devote a large part of their day to the project.

7. Choosing the wrong project leadership. Mistakes include identifying two project heads without stating who has authority, and, two, creating a working committee that is too large. A project leadership team with 20 participants is "inertial," Forrester reported. "An informed few can proxy for the many, and only one of the few should clearly have final decision-making power," Forrester reports.

8. Distrusting the managers to whom tasks have been delegated. Another way to put it: avoid micromanaging. Do not use steering committees to delve into technical details, for example. That only alienates managers and takes valuable time away from making decisions that are needed to advance the project, Forrester concludes.

9. Jumping into the “D” of “R&D” without enough “R.” Schedule pressures, and an enthusiastic technical staff, tend to take away time for properly scoping out a project and learning exactly what is being asked. Build in investigative time into schedules, Forrester advises.

10. Suppressing bad news. Some organizational cultures view problems as embarrassing. That means the staff is reluctant to point out problems, even it means bigger problems will occur later in the project cycle. Discuss problems as they occur, Forrester recommends.

The FBI's effort to upgrade its computer systems -- a program that has had numerous missteps and failures over the years -- is again under fire. This time the Government Accountability Office concludes in a report released this week that the FBI's new network has major security lapses that leave the system open to hackers both inside the agency and out.

The GAO concluded that the bureau hasn't followed some of the most basic security practices when modernizing its computer networks, a program aimed at allowing agents nationwide to share evidence in investigations and to better manage their own cases.

The list of shortcomings is long, including not installing identity management controls to filter out unauthorized users, encrypting sensitive data, not recording or monitoring who accesses sensitive information, or updating software patches on a timely basis to protect the system from the latest viruses and security holes. "Taken collectively, these weaknesses place sensitive information transmitted on the network at risk of unauthorized disclosure or modification, and could result in a disruption of service, increasing the bureau’s vulnerability to insider threats," the GAO concluded in what can arguably be characterized as an understatement.

The unnamed network is part of the FBI's troubled Trilogy program to upgrade the antiquated information networks at the bureau. Trilogy has had a series of setbacks, cost overruns, delays and failures, the most notable of which was the failure of a system to allow agents to share evidence and other information on cases they were investigating, called the Virtual Case File. The FBI killed the system after spending $170 million on the project. Trilogy ran into serious enough problems that Congress began holding hearings on the progress of the modernization effort as early as 2002.

Hat tip: Wired

If you needed a reminder that government information technology does have an impact on the daily lives of the public, consider what has been playing out in Maine for the past two plus years.

The state, along with contractor CNSI, have spent the past six years building a computer system to process Medicaid claims that doctors, hospitals and other health care clinicians submit for payment. Right from the start, however, the system had numerous software problems, which caused many Medicaid recipients to not receive health care and which delayed payments to health care providers, creating serious financial problems for many. After spending more than $70 million on the system (the original cost was $15 million), the state decided to kill the system and seek to outsource the claims processing. (I wrote about the problems in a feature for CIO Magazine last year.)

Despite the system's improved performance and the fact that Maine plans to outsource the work, Maine health care providers, advocacy groups and citizens still are expressing anger, as a sharply worded editorial that appeared last week in the Kennebec Journal illustrates. "Discredited functionaries" is how the editors described the state's public managers. That editorial drew an equally strong response today from the commissioner of Maine's Department of Health and Human Services. "Perhaps hyperbole has no bounds," Commissioner Brenda Harvey fired back.

Maine plans to contract with a private firm to manage the claims processing work and to provide a system that can expand as the federal government's Medicaid program demands increase. That contract should be awarded in about three more years.

A group of former and current State Department employees is calling into question the merits of an award-winning computer system designed to speed up the process of conducting security clearances, calling it "garbage in, garbage out."

The Concerned Foreign Service Officers, a group formed in 2005 by State Department employees who are concerned about abuses of the security clearance process, issued a press release Friday questioning the recognition that the State Department's Bureau of Diplomatic Security has received for developing a computer system that decreases the time it takes to issue security clearances. The security-clearance process has strained under an increased demand since 9/11, with the backlog of security clearances remaining at more than 100,000 and taking more than a year to complete.

State Department employee Donald Reid "recently received an IT Leadership Award from Government Computer News, for innovation in the use of automation to facilitate the security clearance process," according to the press release. The security clearance system also received the Guardian Award from the Office of Personnel Management, according to the press release.

The Concerned Foreign Service Officers group agrees the system is an undeniable improvement over the paper-based system and reduces the time necessary to gather information for security clearance adjudications, but:

It is important to understand that the speed of information gathering is the ... least important aspect of the security clearance process. Two other aspects are considerably more important: the quality of investigative reports and the quality of adjudications. Like medicine or science, a security clearance adjudication is a qualitative function. To focus on the speed of the process is a bit like focusing on speed of intake, rather than diagnostic or treatment abilities, in a hospital.

In its press release, Concerned Foreign Service Officers congratulate the Bureau of Diplomatic Security on increasing the speed of its security clearance process, but "we only lament that the primary result of this improvement is that DS now makes bad determinations faster."

The State Department public affairs office declined to comment.

The Customs and Border Protection is not properly tracking whether foreign goods sent into the United States under a federal program allowing the goods to delay payment of duties and inspection are in fact being assessed duties and being inspected at another port, according to a Government Accountability Office report.

Under the in-bond system, some foreign companies sending imports to the United States are allowed to delay duties and inspections until the goods reach another port. Many of these imports include goods that eventually will leave the United States bound for another country. The program was designed to keep goods moving quickly through crowded U.S. seaports, where infrastructures have not been able to keep up with a dramatic increase in imports. (The value of U.S. imports more than doubled in an eight-year period, from $881 billion in fiscal 1998 to an estimated $1.82 trillion in fiscal 2006, GAO reported.)

Under the in-bond program, the CBP must charge duties and inspect the goods if necessary when the goods reach another, less busy, port, reconciling the charges and inspections with the importer's in-bond documents in a computer system. But many of the documents are not reconciled, GAO charged. At the Newark, N.J., port, more than three-quarters of all in-bond shipments were unreconciled, and some ports, such as Los Angeles, the port with the largest amount of in-bond shipments, could not provide any data on how many in-bond shipments remained unreconciled.

GAO concluded that the large percentage of in-bond shipments that are not reconciled means a loss of revenue, a potential for fraudulent goods being dumped on the U.S. market and ultimately a security risk. "Lack of accurate information on the value of in-bond cargo prevents CBP from accurately determining the extent of any lost revenue," GAO reported.

CBP agents use information systems to reconcile the documents. Agents will be able to use a new trade system called the Automated Commercial Environment (ACE) is expected to improve the process but is not expected to be ready for more than eight years, GAO reports. Development of the system, however, has been plagued by numerous problems. In its report, GAO wrote "that CBP faces long-standing management challenges and new risks associated with the development of ACE."

A record number of taxpayers electronically filed their individual tax returns with the Internal Revenue Service for the 2007 tax filing season, the agency reported Tuesday.

The 76.7 million e-filed returns the IRS accepted through May 4 topped the 73.2 million electronically-filed returns received for all of 2006, with most of the increase coming in March and April, the IRS reported.

The agency's Web site, IRS.gov, also hit a record, receiving 140 million visits. A record 22 million people filed electronically from their home computers.

“E-file and our other electronic services helped us deliver a strong filing season for the nation’s taxpayers,” IRS Acting Commissioner Kevin Brown said in a statement. “Again this year, millions of additional taxpayers gave up paper tax returns to file electronically."

According to the IRS, the 2007 tax season saw a surge in electronic filing among last-minute filers, a group that has traditionally sent in paper returns. During the week that included this year’s tax-filing deadline (April 14 to 20), the number of electronically-filed returns received by the IRS jumped 35 percent over the comparable week last year. The overall number of returns (paper and electronic) received during that week rose only 12 percent, the IRS said.

Despite the positive numbers, the popularity of the Free File program -- an alliance of companies that offer free return preparation and electronic filing on their Web sites to eligible taxpayers -- for electronically filing federal income-tax returns continues to decline, according to the Government Accountability Office. According to an April GAO report, taxpayers' use of the Free File program declined 5.5 percent from the previous year. Free File, now 5 years old, is available to taxpayers with an adjusted gross income of up to $52,000 or about 70 percent of U.S. individuals.

Introducing new technology into old ways of doing business can be challenging at times. That's what the U.S. Census Bureau is finding out, Government Executive observed during a recent trip to a site where the bureau is testing wireless computers.

In Fayetteville, N.C., dozens of temporary census employees are testing how new handheld computers perform in the field to verify addresses and add new ones. During a training session that the bureau held last week, many of the trainees' handhelds froze as they tried to input or download information. The handhelds, which are linked to census databases, upload and download information via satellite and require about 10 to 15 seconds to respond to requests. During that time, trainees, thinking the computer was not operating properly, continued to tap their stylists on the touch-sensitive screens. That caused the handhelds to freeze.

The bureau is testing the use of the handheld computers in a nine-county region in and around Fayetteville as well as in Stockton, Calif., to check how well the devices help employees verify addresses. The handhelds, which are outfitted with GPS location devices, will replace the paper, pencils and maps that enumerators carried around during the prior censuses to locate addresses and to record answers from individuals who had not mailed in their census forms. Last year, Harris Corp. won the $600 million contract to supply the handhelds. Census hopes the handhelds will reduce costs for the decennial census (the 2010 census is estimated to cost $11.3 billion compared with $6.6 billion in 2000) and make the census more accurate.

This dress rehersal was conducted to uncover problems. And it did. Other problems with the handhelds included the device's fingerprint scanner, a security feature that prevents anyone other than the enumerator from accessing data on the handheld. But Beatrice Wolff, a 70-year-old retiree from Fayetteville, found the scanner didn't always work for her. She said the handheld's fingerprint reader repeatedly failed to recognize her fingerprint when she would try to turn the handheld on, causing the automatic shutdown feature to kick in. That denied Wolff from accessing her computer for 15 minutes before she could try again. "This [handheld] is giving me a lot of problems," she said.

Monique Moya, a crew leader overseeing eight people canvassing the Ft. Bragg military base in Fayetteville, said at any time as many as 25 percent of the eight listers she supervises were experiencing some kind of problem with the handheld computers in the field. Most problems were software related, she said, or because the satellite communications didn't provide enough bandwidth. "When the handhelds are working, it's great," she said. "But with this being the first rollout, all the bugs are showing up."

Harris will spend the next year trying to iron out those bugs. A Harris technician at the bureau's Fayetteville office, who was busy fielding calls trying to solve problems with the handhelds, declined an interview. The bureau and Harris have less than two years to fix the bugs before the bureau begins to verify and add new addresses nationwide.

When it comes to IT projects frequently failing, Colorado is no exception. The state's troubled $223 million welfare benefit system is just one example.

But the state legislature is trying to do something about it, according to an article posted by the Rocky Mountain News. The Colorado Senate passed Senate Bill 254 abolishing the Colorado Commission on Information Management, which was compromised of lawmakers, private-sector experts and department heads who oversaw IT projects.

Taking over those duties will be the Colorado Governor's Office of Innovation and Technology, comprised of much of the same individuals: tech specialists and department heads, who will draw "on outside experts," according to the article. "The idea is for the governor's respected Chief Information Officer Michael Locatis to forge better collaboration and expertise-sharing among information technology teams now scattered across 20 agencies, said Rep. Bernie Buescher, D-Grand Junction," the newspaper reports. "An executive with strong private- and public-sector IT expertise, Locatis won praise as Denver's technology czar for forging the city's fragmented technology offices into a strong team."

The Rocky Mountain News quotes Buescher:

This is an effort to say: Let's get our very best minds together. Let's concentrate our effort. Let's make sure that when we do a new technology program that it's driven from within one department.

Is creating another office to oversee IT projects enterprisewide the answer for failed technology projects? Or is the key to IT project success a strong central leader? Or is it something else? Let us hear how you feel by clicking the "comment" link below.

The U.S. Census Bureau has not developed an effective computer-training program for the thousands of temporary workers it plans to hire to interview citizens who may not send in census forms for the upcoming 2010 census, according to a report released last week by the General Accountability Office.

Census officials plan to outfit an estimated 525,000 enumerators with handheld computers. Census hires enumerators as temporary employees to track down individuals who have not filled out census forms. The enumerators will use the handheld computers to input answers to census questions and then later download the data to Census databases. The handheld computers, provided by contractor Harris Corp., will replace the paper-and-pencil process enumerators have used for decades.

But the GAO warns that the Census Bureau's hiring procedures do not look for candidates who have computer skills. For example, crew leaders, those in charge of supervising enumerators, will be in charge of troubleshooting any problems with the handheld computers. But the Census does not plan to ask candidates for crew leader positions if they have computer experience and skills that would allow them to be effective in fixing any problems that may arise with the handheld computers. The GAO concludes:

The bureau is providing some computer-based training on using the handheld computers for the nonresponse follow-up and address canvassing operations and will include visual aids to enhance training on using the handheld computers. Nonetheless, the bureau’s standardized approach to delivering training, including reading training scripts word-for-word over the course of several days, has remained largely unchanged. The bureau has not evaluated alternate training delivery approaches, such as providing video segments, as has been recommended by us and the [Office of Inspector general].

Federal law prohibits the sale of guns to anyone judged mentally ill, but most states are unable to share mental health records with an FBI computer network that would block the sale of guns to the mentally ill because of privacy laws or state computer systems that are incompatible.

That may change if a long dormant bill in Congress -- revived after the shootings at Virginia Tech -- is passed. The bill would provide $1 billion to states to pay for computer network upgrades and to remove privacy law obstacles, according to an Associated Press article. According to the article:

Privacy laws and lack of technical ability now prevent 28 states from sharing such information with the FBI’s National Instant Criminal Background Check System based in Clarksburg, W.Va., according to a Justice Department report.

“Every one of these records that is not transferred is the record of someone who federal law has said is too dangerous to buy a gun,” said Dennis Henigan, legal director of the Brady Center to Prevent Gun Violence.

Such a system should have prevented Seung-Hui Cho, the gunman who killed 32 people and himself at Virginia Tech, from buying guns. In 2005, Cho was declared mentally ill by a special judge's order, according to a New York Times article.

Automating business processes is supposed to create efficiencies. But for the Wisconsin's Department of Motor Vehicles, a new computer system has resulted in the opposite outcome.

A new system installed in 2004 to reduce the time it takes to receive a license plate and vehicle title has more than doubled the wait time -- from three weeks to seven weeks, according to an article in the River Falls Journal. In addition, the cost of the system ($19 million) also more than doubled what was originally budgeted.

DMV officials now say "adjustments" to the system should drop the wait for license plates and titles to 30 days.

Wisconsin's government computer projects are failing because of poor planning, cost overruns, delays and a lack of oversight, a report released by a state representative shows.

Sue Jeskewitz, R, Menomonee Falls, who oversaw the Legislative Audit Bureau report, "says we need project managers, for accountability, and the state should think about looking into new contractors. Jeskewitz says the Department of Administration, which was too busy with its own problems to address other problems, has lost credibility," according to an article on Wisconsin Radio Network's web site.

Jeskewitz plans to hold a hearing May 2 of the Joint Legislative Audit Committee to review the report's results.

The Internal Revenue Service encourages as many taxpayers to file electronically as possible. They may have got their wish, but now they have another problem.

So many taxpayers submitted their returns electronically on April 16, the deadline for having your taxes filed, that the servers at Intuit Inc., which processes the electronic tax returns for the IRS, became overloaded and slowed the filing of e-returns by hours, according to an Associated Press article. The delays may have caused many taxpayers to have missed the filing deadline.

Under normal working conditions, it takes a few minutes for an electronic tax return to complete its submission using TurboTax, according to the AP. But by late Monday, it was taking hours.

No word yet if IRS officials will grant amnesty from late penalty charges for those filers who missed the deadline because of the overloaded servers. But the lesson here, says Harry Pforzheimer, an Intuit spokesman, “Don’t wait until the last minute is the moral of the story."

But some IT experts may rebut, quoting the Boy Scout motto: Be prepared.

Maine's Department of Health and Human Services will pay another $7 million on a failed Medicaid claims processing system before it can kill the project, according to an article in Maine's Times Record.

The total cost of the claims processing system could surpass $70 million, close to five times the amount Maine agreed to pay in 2001, when it awarded a $15 million contract to CNSI to develop the system. When the system was switched on in January 2005, it immediately began to have processing problems. IT program management experts blamed the state for not following best practices for project management, as reported by CIO Magazine.

Maine decided earlier this year to scrap the system in favor of privatizing the claims processing system.

This post was written by Karen Rutzick, staff correspondent for Government Executive Magazine.

NASA is retooling their Web site, and they’re doing their homework first. NASA Internet Services Manager Brian Dunbar is conducting extensive interviews with Web site users, including the media, such as this reporter.

Dunbar says they’re “kicking off a full re-design of the site,” including the media section. He wanted to know how useful the current offerings were to reporters, what else could be added, how often sections were used, and other information.

NASA’s reaching out to members of the general public for feedback, too. A customer satisfaction survey asks Web site perusers to weigh in on the ease of reading the site, the clarity of its organization, search results and trustworthiness.

Software design experts routinely warn that one of the pitfalls of system and web development is not to ask users for input. An article in The New York Times last month (which ran with the headline: "How to Improve It? Ask Those Who Use It") discussed the advantages of so-called "citizen product design." Here're some tips on interaction design offered by Asktog.com, operated by the usability designers Neilson Norman Group.