999 What's your emergency: Mega millions Met call handling IT muckup? A major delay to the Metropolitan Police's new command-and-control system responsible for handling 999 calls will cost the tax payer £25m, it has emerged. The £90m contract to overhaul the Met's 30-year-old command-and-control system was meant to go live last year after Lockheed and Northrop Grumman won the 17-year " …
Please somebody explain.
Government pays for product
company paid to deliver product fails to deliver on time
Government pays more money
Why are the companies failing to meet their target not carrying the weight of the failure ?
All of the above comments actually boil down to what I Am Spartacus said at the beginning.
The other thing I wonder about is just how much of this 'flaccid' contract writing is fostered by the big operators in the first place with them offering platitudes about changing specifications being able to be covered rather than requiring a tight specification to start with.
It all comes down to the product specification, brief and contract.
Also the product milestones and development methodology. No company would take on a contract which states - we'll pay you at the end and you must deliver exactly what we had in our minds and what we subsequently thought of.
Similarly no customer would submit a specification and brief for a project this size that included every screenshot, process flow, text box size, colour etc, that contains everything exactly as is required.
Somewhere in between there will be an expectation gap, and how this is handled costs the money and/or the time. Eventually they can hit the point where they can't afford the future milestones but also can't afford to throw away the money already spent on a system that isn't finished and can't be used. Therefore if they abandon you hear stories of xyz has thrown away x million on another failed gov it project and if they carry on you hear stories of xyz has overrun budget by x million and taken x years more than expected.
The suppliers who are willing to take on these projects are few and far between so the usual suspects keep cropping up, they also don't often have a fixed workforce as they never know when the next massive project will turn up and so they rely on a contracted workforce working at different qualities and using different methods.
Overall it's not surprising there is so many failures and issues. However a start would be a brilliant brief with a rock solid technical and functional specification that delivers fully working milestones that can work independently - starting with the core requirements and then having further milestones bolted on as modules, black-box style.
> However a start would be a brilliant brief with a rock solid technical and functional specification
> that delivers fully working milestones that can work independently
Mmm, but to have the capability to really deliver that specification you probably need a serious in house IT capability, including the experience of running the current system with all the realworld problems that throws up.
The requirements are usually poorly defined and the suppliers won't commit to a fixed price for the job because of that - for the same reason that a builder won't give you a quote for 'a sort of big house thing, probably some bathrooms and stuff in it'.
They contract then on time and materials terms - an hourly rate if you like.
What tends to happen is at the end of the dev cycle the thing doesn't work right. It does however conform perfectly with the poorly written requirements and so the supplier wants paying. Fixing it then requires another cycle of requirements, development and testing. Excitingly at that point the supplier is often replaced with a new one who spends the first 3 months trying to work out what the old supplier actually did and learning the business problem from scratch.
After nearly 30 years of this on both sides of the fence the only answer I can come up with - the thing that demonstrably works - is to directly employ your development team. Make solving the business problem a thing that benefits their prospects. Make urgent fixes just part of their employee responsibility rather than something the supplier will milk. It's unfashionable and the MBAs get stroppy, but it works and it's cheaper than every alternative as far as I can tell.
I really don't know, but I'd love to get a government contract...
You get the contract due to the best spec/quote/delivery time scale (okay, so that's normal)
Then you fail to deliver on time, fail to meet the spec, over run the budget, and they not only don't chuck you out for breach of contract and black list you from future bids, they throw more money at you!
Kerching!
Government contracts are just the same as the private sector ones.
The private sector do this all the time but are not compelled to make it public.
So really you want a private sector contract, you get just as much money but your reputation is still good afterwards.
Ah, I see you don't deal with tenders often then.
It generally works like this....
Stupid, flawed tender is created. It'll be stupidly detailed in some areas, normally the areas where flexibility and choice would be useful, and often demand things that are not practical, cost a fortune or absolutely don't make sense. Then in the areas where some detail would REALLY help define what it is you're actually supposed to be doing, it'll be incredibly vague.
Tender is sent out, with limited way to change this, various questions are raised, unhelpful answers are received (such as "please quote as per the tender"), or further incomplete or conflicting answers are given.
Tender is won by whomever (normally the lowest bidder, or the company they really wanted to do it based on making sure they meet whatever criteria were set).
Tender is then largely thrown away and ignored, and a million change controls, specification changes and alterations ensue. The winning party gets to revise quotes a million times and change prices, the original party wanting the work done repeat the vague/pointless/detailed cycle, and so on and so forth.
Some time later something not resembling the original tender is produced, does not work because of all the aforementioned issues, and when a problem arises the winning party can point to the string of change controls and the audit trail as "you changed the specification" as the answer to all questions on timescale, functionality and so on, and anything where "the specification was wrong" arises will be dealt with by further change control and greasing of palms.
It's a slow, inefficient model, doesn't provide anyone with value for money (except perhaps the party winning the tender) and is so badly flawed it is unreal, but the winning parties keep winning with the current setup, and those who are tendering know it's shite but equally can tick all the compliance boxes required.
Missing from that overview are many pointless meetings and conference calls, none of which answer any actual issues, but cause further problems, and a whole heap of other red tape and nonsense that serves nobody well.
Simple. Gov buys in to your companies spin as it's similar to theirs. At that point both Gov and IT spin similar crap and both know that us working folk pay our good pennies into this whether we good folk like it or not.
At worst one bloke, literally one individual, takes the rattle and resigns on a huge pension plan (plus the bonus £££ awaded by winning the contract) due to a "gaff".
He will continue on not as an MP or CEO but as a consultant to his other Cholmondley-Warner type Eton pals earning more than ever. His postion is great and won't ever rat out anybody else. Harvard Rules or something suchlike.
This post has been deleted by its author
"Without wanting to sound too much like a UK prime minister (most of them anyway...), I can't think of any of my problems that an indiscriminate airstrike wouldn't improve to some degree."
Well I don't think it would do much for the parking problems caused in my road by my idiot neighbour...
Although I'd love to see it anyway!
The ideology that a for-profit company can provide a service more efficiently than an in-house organisation suits "small government" conservatives to the ground. But a business contract is always going to be about the supplier trying to maximise profit and minimise cost to themselves. It's a business deal, not a favour. So they are not going to provide better than they've agreed to provide, even if that's in the spirit of the agreement.
And statements like; enabled for multi channel public interaction delivering improved situational awareness, predictive analytics and data sharing, just reek of meaningless bullshit.
Sorry, but i'm not sure i agree with that. I'm not particularly in favour of a "small" or "big" government, but it doesn't always make sense to do these things in-house. it depends entirely on how many times you are planning to build a system like this and how you want to attract the staff to work on it in the firstplace and what you plan to do for them at the end of the project.
In fairness to the police, their job is use the system, not build it themselves. It's a big technical thing that they need and the amount of expertise required is significant - if they had to recruit all those people to build it there would be a huge lead time in finding and setting up those teams followed by all the same problems in providing the right specification (which yes, is crucial). Then you need to get rid of most of those people once the thing is built (because you need more people to build it than maintain it) - the admin alone would be monstrous.
So unless the government want to set up an IT infrastructure building business (which they could do, but would probably still have to compete with the capitas of the world) does it really make sense to try and do this directly, when it's not the Met's area of expertise and it is only an infrequent requirement?
"it doesn't always make sense to do these things in-house. it depends entirely on how many times you are planning to build a system like this and how you want to attract the staff to work on it in the firstplace and what you plan to do for them at the end of the project."
Development spend to launch is 10% of whole life spend.
You'll find plenty to keep those developers busy with until the day you turn the system off.
was a total cock-up when it first went on line on top of all the turf wars.
But their PR system is smooth ... " Office of Counter Terrorism (OCT) Intelligence Unit also staff the New York State Intelligence Centre (NYSIC). Managed by the New York State Police, the NYSIC serves as the State's Fusion Centre, bringing together federal, state and local agencies to analyse and share information related to terrorism and other crimes."
Meanwhile, the FBI is getting it's security theatre to produce 'terrorists' for the New York State Intelligence Centre to look good.
Has anyone ever found out why the RAF helicopters are flying nightly sorties sniffing RF signals over London and the fixed wing MET aircraft - all based at RAF Northolt - a CIA rendition flight centre - are flying every night sniffing cell signals from the south of England in to Scotland?
Also "the Met's 30-year-old command-and-control system"
In 1989, I worked with the Unisys guys (no gals IIRC) converting the old Sperry 1100-based system onto flash new Unisys 2200 hardware. The system was certainly already more than four years old then, so I'd put the system age at more likely 35 to 40 years than a sprightly and youthful 30.
I'd expect that the software was based on systems designed and written by PRC, a US-based Sperry partner who specialised in "Computer Aided Despatch" applications on Sperry hardware. I guess it could be used for taxis, but the package and the hardware it needed cost several millions of dollars (US) in the 1980s, so I doubt very much it was affordable by any taxi company.
Mike
" I guess it could be used for taxis, but the package and the hardware it needed cost several millions of dollars (US) in the 1980s, so I doubt very much it was affordable by any taxi company."
FWIW the US Marine Corps used to describe their AV-8A/C Harriers as "bomb taxis" as essentially they were originally used as flying artillery
The UK ones were a lot more capable with better targeting gear
Since there are at least two well-known public safety/incident management systems available for purchase off-the-shelf (yes, with a pretty heavy support cost, and some customisation required... but it's proven technology), what makes the Met different to the British Transport Police, which uses one of these systems?
Not to mention all the other police forces and fire and ambulance services that use these products around the world. The Philippines alone covers 100 million people across a number of public safety agencies. And it doesn't take £25m to implement either solution that I'm aware of (not for the likely size of the implementation). Maybe Northrop/Lockheed are just going to rebadge one of these and slap a bit of who-knows-what on top.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
