As I was driving to the National Motorcycle Museum to give a presentation on archiving this morning, I listened to a constable from a Welsh police force on Radio 4, backing up the soon-to-be-announced (or has it been already?) statement that there was too much bureaucracy in the UK police force. For anyone who has difficulty (as I do) spelling bureaucracy, it is helpful to remember that the first part is "bureau", i.e. "desk"—which is exactly where our local bobbies seem to spend much of their time.

The constable was explaining the overheads involved in stopping a group of individuals and asking whether or not they had any knowledge of an incident (say, a punch-up round the corner on a weekend evening). Each conversation needs to be registered on a form, which takes 10 minutes to fill in—one can only imagine the number of such conversations that take place in central Cardiff on a Saturday night.

Now, of course we need to balance any statements about "overheads" with equal remarks about protecting the rights of the individuals that are stopped and questioned—but that doesn't seem to be the point here. Rather, one is left with a feeling of, if police officers have to talk to the public for anything other than passing the time of day, there will inevitably be more time spent documenting the conversation than holding it in the first place.

This point was reinforced by an IT professional at the archiving event, who happened to be from a city-based police force. I have to ask—is this as far as technology has come, if we still need our police officers to be filling out sheets of paper in triplicate? I had a bit of a chat about the alternatives to form filling—voice recognition, for example. To support the average copper, with the average regional accent, such technologies are still not yet ready for prime time, I am informed. "Police officers need technology to just work—not to sit through hours of training the software for it to understand them," I was told. This seems such an utterly fair statement given the 20 years of innovation we have seen since I kicked off my career—that technology should just work—so why is it not true?

The answer, I believe, lies in the gap between technology companies and systems integrators. Some of the things that end-users see as fundamental are not baked into the base software, but are left to partners to implement: an example in the voice recognition context would be the level of training required to support it. Surely there are mechanisms that can integrate with how people work today—passive voice recorders, local accent filters, document scanning etc—that could be used to make such technologies "just work" when used by people on the front line of the business?

I'm not totally despondent—but it's situations like this that serve as a constant reminder of, even with all the innovation going on, just how far we have to go with technology before we work out how to close the gap between the wonderfully clever capabilities it offers, and the real needs of its users.