$145m project collapse: AFP fails to upgrade 18-year-old case management system


blog As regular readers of Delimiter will know, Australia’s police forces have not precisely covered themselves in glory when it comes to upgrading their ageing IT systems. Victoria Police’s IT systems are so bad they are contributing to fatalities, NSW Police is still struggling to upgrade its own 15 year case management platform, and there are even cases where police forces have been busted allegedly pirating software. Not precisely a good look. Now it’s time for an IT bungle at the Australian Federal Police. iTNews reports (we recommend you click here for the full article):

“The Australian Federal Police has cancelled its $145 million investigation systems replacement and sacked integrator Elbit Systems after deciding the project had become irretrievable due to time, cost and functionality shortcomings.”

I would say this is quite an extraordinary move, but of course it’s basically situation normal for major public sector IT projects. One wonders whether someone — anyone — will be held accountable for this mess. One would hope so, given the fact that the AFP has apparently been trying to do something about this platform since at least 2007. Maybe it will succeed with some modern upgraded by about 2020 — if it pushes on forward with the job. And let’s not forget, these are the people we are trusting with Australians’ metadata. Wonderful.


  1. Is the problem more around the inability to get acceptance of requirements? Onerous requirements for integration? Antiquated integration points?

    When an 18yo system is still in operation there’s big questions as to *why* the failure occurred? Purely technical? Purely people? Purely systemic? Where the combination of the three combined?

    From my time in bigger orgs some of the problems boil down to the following:
    – must do everything the old thing did (even look the same)
    – cannot reduce functionality (even if the requirements are questionable)
    – must be delivered in one hit (no agile methodology of continual improvement allowed, sometimes this is understood)
    – efficiency gains aren’t allowed due to diminished fiefdoms

    • Unpleasantly familiar list. I would only add:
      – In addition to doing everything the old thing did, must also incorporate 25 new features that the business has developed workarounds for over the life of the previous product.

  2. Sorry I’m not clear from the commentary, had they actually spent $145m already or did they stop it before it went that far?

      • That’s what I was hoping someone could answer for me. It may have been a $145m budget but if the project was stopped at $1m then someone had the guts to realise early that they hadn’t done their prelim work well enough. If they got all the way to $145m and then found they had nothing I’d be thinking someone should be investigated for criminal negligence

Comments are closed.