Top 5 government IT disasters besides

Top 5 government IT disasters besides

Summary: The government program generally known as Obamacare is the poster child for poor government IT work. There have been others though.


 |  Image 6 of 7

  • Thumbnail 1
  • Thumbnail 2
  • Thumbnail 3
  • Thumbnail 4
  • Thumbnail 5
  • Thumbnail 6
  • Thumbnail 7
  • The FAA Advanced Automation System

    The Federal Aviation Administration's Advanced Automation System (AAS) project was supposed to provide a complete overhaul of the nation's major air traffic control computer systems: new tools and displays for controllers to improved communication equipment and a revamped core computer network. In the end, the FAA decided that $1.5 billion worth of hardware and software out of the $2.6 billion spent was useless.

    The testimony of a GAO official before Congress in 1994 makes clear that it was bad management which caused all the problems:

    AAS's cost and schedule problems have resulted from several technical and managerial factors. First, FAA and IBM's development and implementation plan, including cost and schedule estimates, was overly ambitious given the highly demanding requirements and the complex software architecture for this system. Second, FAA did not provide adequate oversight of IBM's performance, especially during the initial development of the key ISSS component. As a result, IBM's lack of progress did not always surface in a timely manner. Third, FAA was indecisive in resolving some issues about basic requirements, such as the format of new electronic flight data strips to be used by controllers. In our opinion, the above factors — not inadequate funding or federal procurement rules, as contended by some proponents of an air traffic control corporation-- have caused the AAS’s problems.

    AAS is another in a long line of projects that underscore what IT is up against in trying to work in government: The procurement process took far too long and was far too political. The new system was supposed to be magically powerful compared to the old one, so expectations for it ran out of control. It also seems that IBM was as unrealistic in the design as anyone and didn't execute well.

  • NHS Connecting For Health

    NHS Connecting for Health, launched in 2002, was supposed to "...[deliver] the NHS National Programme for IT (NPfIT), an initiative by the Department of Health in England to move the National Health Service (NHS) in England towards a single, centrally-mandated electronic care record for patients and to connect 30,000 General practitioners to 300 hospitals, providing secure and audited access to these records by authorized health professionals."

    Sounds ambitious and wouldn't it be cool if it worked! It didn't. Go to the Connecting For Health web site and you see a big banner at the top that says "NHS Connecting for Health ceased to exist on 31st March 2013. This website is therefore not being updated." It refers users to a different site.

    The BBC reported that "[a] report by the influential Public Accounts Committee (PAC) concluded an attempt to upgrade NHS computer systems in England ended up becoming one of the "worst and most expensive contracting fiascos" in public sector history." The most recent estimate of money wasted is £9.8bn, but even that doesn't account for everything.

  • Dishonorable Mentions

    • Déjà Vu All Over Again: California's DMV IT Project Canceled — The state of California's Department of Motor Vehicles (DMV) is 0 for 2 now on IT modernization projects. In 1994 it canceled a project begun in 1987 after spending $44 million. Earlier this year, six years and $134 million into a second go at it, the state canceled it too. Doesn't anyone in California know anything about computers?
    • - The System for Award Management — SAM, which is designed to integrate three acquisition data systems that store and make available information about contractors, went online in July 2012 and was taken off-line days later due to performance issues. It's up and running, but some claim it's still not running right.
    • Federal Protective Service Risk Assessment and Management Program (RAMP) — program useless, massive cost overruns - The Department of Homeland Security (DHS)'s Federal Protective Service (FPS) Risk Assessment and Management Program (RAMP) — Enough Acronyms For You? (EAFU?) — is designed to facilitate the FPS's mission of securing Federal facilities and ensuring occupant and visitor safety. A report by DHS's CIO says that "RAMP’s development and deployment has been delayed for two years, and its Life Cycle Cost Estimate (LCCE) has grown from an initial estimate of $15.9M in 2008 to $183M in draft 2011 LCCE. RAMP’s LCCE growth is attributable to poor initial estimates, scope increases, reprioritization of capabilities, unplanned fixes to technical issues, and extended system lifespan."

Topics: Government, Government US, Government UK

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Related Stories


Log in or register to join the discussion
  • And PRIVATE IT projects never fail?

    Contrary to what the anti-government people want you to believe, it is not only government projects that fail badly. Private companies have disasters also, but they are generally only known to the employees of those companies and the insiders in their industry. I worked on one in the 1980s in the trucking industry, and I remember many sleepless third shifts (AFTER working the first and second shifts) and shopping carts full of memory dumps. The company recovered, as the big ones usually do (having a big cash reserve forgives many errors), but not all of the employees' marriages did (I was lucky).

    The causes of IT failures include over-optimism, executive override of technical warnings (analogous to the Challenger disaster, and remember that was in the private company that built the hardware), miscommunication between teams, lack of prior planning (in one case, a holding company owned two subsidiaries with incompatible systems for ten years with no research done on "what if" they merged, then announced the merger and pushed it through in 18 months), mission creep (which may be MORE of a problem in the private sector, since requirements changes do not require changes to laws), and basic human ego, carelessness, impatience, etc.

    Certainly, government projects tend to fail bigger than private ones, and affect more people when they do. They are also, as I mentioned, harder to cover up, and more likely therefore to be widely known outside the specific industry involved. But the same political and individual failings are present in all human endeavors. Errare humanum est.
  • It's not only the USA

    The near total collapse this last weekend of a relatively new National Air Traffic Services system in the uk was "The fault occurred when the Nats computer system was making the switchover from the quieter night-time mode to the busier daytime setup." ! ??

    BBC news 7 December 2013
    Thousands of flights have been delayed due to problems at a key air traffic control centre in Swanwick, in Hampshire
    The facility, which handles 5,000 flights every 24 hours, was delayed even before it opened, eventually taking over operations in January 2002 - six years after its planned commissioning date.

    (Air traffic control system is 'not safe', say UK controllers | ZDNet)
    By Tom Espiner | March 2, 2011 -- 16:48 GMT
    Technology being introduced at one of the two major UK air traffic control hubs is "not fit for purpose" and did not adequately handle a breakdown in air traffic communications..
  • Procurement seems to be a common thread

    The process is in desperate need of reform, but it's not going to be at all easy, as there are too many vested interests (a relatively simple chain of command would probably be helpful). I don't think there's any excuse, however, for administrative functions to be outsourced either in government agencies or private enterprises (especially large corporations). It's reasonable to outsource designing and building new systems, but the people running and maintaining them should have their careers invested in the organizations they're serving, not outside firms.

    Unfortunately, transferring all government system administration and maintenance to civil service and military personnel would make life awkward in the short term for small government advocates in Congress, but it should be done anyway; government contractors don't show up in the official numbers (maybe they should), but it seems to me that "big government" manned by contractors is not any more virtuous or economical than "big government" manned by regular employees.
    John L. Ries
  • I wouldn't exactly call a disaster

    It took, what, about two months to shake off the bulk of its startup problems? For a project from scratch of this size, that's actually very, very good (go Google up "Microsoft 365 ongoing problems" or "Northrop Grumman Virginia problems" or "IBM Bridgestone".)
    • The ACA system is a disaster, but not for the reasons that you think

      The vision is bad, the marketplace could have been better, but this is certainly bad, and it could get really bad.

      1) Lack of transparency -> crashing from load. The politicians knew that the cost savings that the POTUS touted of $2500/family weren't gonna happen. The politicians did not want a singular journalists or researcher to have access to all of the data for all states, so they forced people to create accounts and start the sign-up process...just to browse. Most people went to browse and get an idea of what they would have to shell out if they are forced into the exchange and overloaded the new system and killed it. The HHS department has not and will not estimate and publish average/expected costs because the population won't stand for it. Now what will happen is journalists/researchers from all the states participating in the exchange will put data together on the side, and the facts will be swept under the carpet and called lies. This will further erode faith in the government system.

      2) Data security/integrity issue from the beginning due to a rushed project and requirement changes. As a part of the process of rolling the system out they granted themselves exceptions to government security standards. Some of the obvious holes have been closed...but not all. As a part of the 'fixes' some data was scrambled. Last statistics I saw is that about 80% of the applications went through to the insurance companies without errors, so that means there is a 20% error rate in the data handling/processing. That may be better now. But given the high-profile website that people put their SSNs into and all sorts of personal information this site will be owned by identity thieves....totally not cool. This could affect us all and force new multifactor identity systems to be put into place for us all as an unfunded government mandate which will have a financial domino effect to the cost of doing business.

      3) Requirements. What problem are we solving? Everyone focuses on the cost of health care, but this entire act is not focused on anything but the cost of insurance. We're not tackling the biggest part of the problem to begin fact we're making it much much worse. The cost of providing health care is skyrocketing and we're not making the health care system more efficient, or more effective. We're adding steps/processing/regulations like people without a plan on the end goal which should be reducing the overhead of providing health care and services. Now the doctors, nurses, and other professionals will have to hire more administrative staffing and fill out more paperwork to do the same things as before. We need to focus on enabling our health care professionals to be the most efficient/effective that they can possibly be. Let me put this another way, we can either cure the patients disease or treat the symptoms, and in this case we're only treating one set of symptoms and creating further complications that will require treatment as well. Fix the core issues with the high costs of health care and the prices will come down.

      The issues above are a pick list of a larger list that I have been considering for some time now.

      On the more positive side where one could make a lot of money is constructing a standardized, secure, durable, scalable, and cheap method for an adaptable multifactor authentication system for everyone in the US along with a distribution method. I'm sure there is a sales person out there that will love this plug. :)
  • These are not the biggest disasters

    The Queensland State Government's payroll system, where it contracted IBM, that is the biggest one. It got rolled out a few years ago with no roll back plan, took ages to fix the bugs, and the political fallout is still happening.

    The mucked up every phase of the project, and still went to implement it. So it'd have to 'win' on the number of bad decisions alone!

    It is the biggest one not just because of the cost of the project itself, but the consequences of its failure. This system would be the payroll tool for all departments, replacing the one each department had built itself. All employees were not paid properly - not just for one pay cycle, but every pay cycle for at least the first year. This meant bills didn't get paid properly, affecting business large and small throughout the entire state.
  • two more

    ln Ontario Canada the provincial government spent $billions to not have a functioning electronic health records system. California spent $billions to not have a state court house computer network.