Auditors blast Blighty cops over binned multi-million pound IT project A new report has been highly critical of a failed £15m IT project for Surrey Police that was eventually binned by the cops. Auditors Grant Thornton published the report today on behalf of the government's Audit Commission, which said the Surrey Integrated Reporting Enterprise Network (SIREN) project had been “poorly managed …
Agile is a fairly poor way of designing very large scale projects. See also the big IDS benefits thing. It's great if you are a start-up or if you are a software company, but when the project gets very big you start to run into architectural concerns that you need to know about from the start and situations where refactoring the code to integrate a new requirement is a non-trivial task.
That doesn't stop the practical day-to-day process stuff from being useful- standups, iterations, sprints, kanban boards and the rest- but as regards the big picture, being able to act on architectural requirements from the start can make a big difference to your development time and the reliability of your software but the concept of doing anything ahead of time doesn't seem to fit into the pure Agile approach.
From my experience with public authorities and their associates Agile means" we don't know what we're doing and we're making it as we go along". The methodology is then used during the project as a blame management tool and nothing ever gets done except lots of money wasted on consultant to tell the customer the bleeding obvious things that anyone vaguely competent would already know.
I designed and built my first submersible ROV over a two month period when I was so stoned I lost weight from lack of appetite. It took the better part of a year before I got caught using it to dredge up coins from the fountain at the university baseball park*, which was what I had created it for to begin with.
So I've got no sympathy for anyone who never got anything useful done while stoned. They were obviously half-assing their attempts.
*This was long before such money was donated to good causes. This money was just dumped into the ball teams general fund, which was bringing in more than $17m annually. So I felt no remorse when the amperage on the power supply on my bench in the lab was adjusted with a pencil I had whittled down to fit the broken shaft on the pot. I replaced that power supply, before I was busted, and I donated the ROV to the school after, but I did list it as a tax deductible donation.
"The auditors said that the PCC was right to axe the SIREN project when he did, but pointed out that the situation never should have come to that in the first place"
Yep they should have used the package that the other forces were using, if they had to adjust how they did things then that's what they should have done!
And don't even get started on why we have 40 odd police forces in this country! Why just in the SW do we have Devon&Cornwall, Avon&Somerset and Dorset ,total waste of money.
But there again the national firearms database was a total cockup and took years to do just for a database that has a few hundred thousand firearms\shotgun cert holders on it.
It doesn't matter what methodology you use - and most are just a variation on common sense - if you don't set deadlines and milestones then the project is sunk. Simple solution to this type of mess - go to suppliers and say "We want this , this and this by this time and you'll get your money when its done. If you over run or under deliver we'll subtract X for each infraction. Now give us your quote.". Sorted. The bullshitters will head for the hills and you'll be left with the companies that can actually deliver.
Sometimes it pays to go to suppliers and ask them what they've got that comes close to your requirements. You can talk to happy customers (if any) and ask them how well it works or where it doesn't. You are also leveraging the lessons learned of multiple existing users as they were incorporated into an off the shelf app.
Not all customer requirements are carved in stone. Sometimes it's cheaper to change a business process to fit a stable software package than it is to tweak the application. And many vendors will consider adding in functions you want in a future release.
Simply amazed, no clause in the contract that says the police can cancel the contract for convenience and pay a sum for that privilege that is less than the final value of the contract (e.g. covers lost profit but not cost of staff). This is a standard in pretty much every public sector contract I have come across.
... about it not being endemic, bollocks I say, this is exactly the right word to use for something all public sector organisations seem to do, pour very rich gravy and see how many companies come flocking for a slurp in a non-risk way through a badly written contract.
Nuke'em from orbit it's the only way to be sure...
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
