Well well well ...
... today I wake up to the news that the UK has decided to scrap its government built, National Health Service (NHS) coronavirus mobile phone track and trace application.
Quelle Surprise ..... this was as certain an end decision, as the news that the sun comes up in the morning. The mobile app had been trialled on the Isle of White in the UK, with over 60,000 downloaders. But guess what, the app was reportedly plagued with technical faults, as well as user privacy concerns.
All this was predictable from the very start:
1). The privacy concerns were apparent before work even started, when the design model of using and storing the data in a centralised server was made known. The idea was that contact-matches were carried out on a remote server. Whilst this had some advantages (particularly in the amount of data epidemiologists would have available), this was also its weakness, as hacking concerns about that data were raised.
2). Due to Apples 'privacy' and code sharing restrictions (as the US government has found, Apple refuse to share development code or password cracking techniques), the NHS app was poor at recognising Apple's iPhones. So while the NHS app recognised about 75% of nearby Android handsets it only spotted around 4% of iPhones. Reportedly the third party code Bluetooth restrictions by Apple, led to this identification failures.
3). The UK's governments incredibly poor record at developing and introducing any software (more on this in a moment).
Why does this matter? Well track and trace is helpful requirement to help prevent a second wave of the coronavirus as lockdown eases. The idea is that mobile apps would log whenever two people have been in close proximity to each other for a certain period of time, so if they are later confirmed to have the coronavirus, anyone who had been in close contact would be advised that they should also get tested and/or self-isolate.
So what are we going to use now? Why the Google and Apple developed apps of course (maybe). These use the decentralised data response, where the contact-matching process is carried out on the handsets themselves, which although not collecting as much data, does respect privacy and possible hacking concerns about data being open to either Governments, and or hackers, to be de-anonymised and misused later.
Now, with this decision, its possible that not even this Google and Apple based app will be used in the UK .... Baroness Harding, who is heading the government response has said that "What we've done in really rigorously testing both our own Covid-19 app and the Google-Apple version is demonstrate that none of them are working sufficiently well enough to be actually reliable to determine whether any of us should self-isolate for two weeks [and] that's true across the world." ... so we wait and see.
In fairness it should be noted that other centralised apps, notably France (the other big promoter of that approach), have also had problems. The European Commission has said that France would face challenges over this approach, including incompatibility with other European countries apps, which are to be based upon the Google-Apple Exposure Notification API, on which they are building their own apps to limit the spread of Covid-19.
Finally, the reason why this expensive U-Turn is no surprise, is the UK governments historic and ongoing spectacular inability to develop and deliver IT projects on time and on budget.
The reasons for this are fairly well known and indeed I posted on it a decade ago, but never acted upon:
1) Lack of Civil Service Skills: The UK civil service hires its Business Analysts and other IT skills from outside as highly overpaid contractors. For reasons that are down to the way it organises itself, its never able to find or train specialists in its own systems, who can model how things currently work, and then detail how any changes would work.
I personally know of a former Cobol developer, who has worked solely in the UK governments employ as an IT contractor for over 20 years without break, at a mind boggling amount of money (we are talking at today's values of well over £1.25 million pounds of tax payers money). He's currently a highly paid 'Business Analyst' contractor (he just took the title instead of developer, and as he knows everyone, was simply re-employed on another government project).
The reasons for this Civil Service failure to produce its own Business Analysts or other IT specialists, is that those who regularly use and become experts in government systems such the benefit delivery or immigration software, work in local offices e.g. Department of Works and Pensions (DWP) ... whereas projects are devised and run by regional office staff who don't regularly use or understand those systems, and who are not impacted by any issues users uncover. See the Universal Credit project.
2) Ministerial Interference: What I.T. requirements that are initially submitted for development, are often changed on a political whim, to meet the ever changing political requirements of government ministers. This causes expensive and dangerous code rewrites ... after a certain point, the code has to be fudged, rather than re-written, because its too expensive to scrap it and start again. So its bad code, and a poor result. See the Universal Credit project.
3) The Project Managers and or IT companies that are contracted to run a project, all rehire the same developers, and analysts from the same small pool of 'good ol'boys,' who have been used on all the other projects. It doesn't matter that many of these have failed on other projects, or frankly couldn't make it in private industry, they just swirl around in the dirty waters of failed or poorly executed government projects over and over again.
4) No firm ownership or decision maker. Heads never roll for failures, as everyone points up the ladder, to a government minister who has usually long moved on. So no one ever feels the career consequences of failure. Consequently no one ever calls a halt to projects that have gone off track. So for example, it must have been apparent some months ago, that with many governments switching to the Google-Apple model (including Germany, Italy and Denmark), for the same reasons that we are now doing, it was likely that we would also be forced to do so. But we didn't stop until millions more pounds of development money had gone down the drain.
... so guess what? The UK Government IT projects all fail, or are poorly delivered, for all the same reasons, involving nearly all the same people and companies every time.
To give you an idea of the problems, the Institute for Government, recently reported that 80% of major government projects are at 'risk of failure' with most being classed as being either in doubt, hindered by problems, or virtually unachievable. This was excluding Brexit related projects, and was based upon the Infrastructure and Projects Authority's annual reports and referred to £455 billion of projects.
So why, given the literally billions of pounds of tax payers money that has been lost to these failures over the last 30 years, does no one reform the system?
Well, again, its inherent in the system that nothing will be ever done:
1) Government Ministers don't want their political hands tied by constraints such as cost or impacts of their late requirement changes.
2) Thousands of Civil Service staff are employed in various Government Projects. its a nice career location, with no public contact, and not strenuous. With no career damage for any failures, so why change? As its almost impossible to sack Civil Servants there is no fear of failure.
3) The governments own Civil Service IT specialists used to be in an organisation called ITSA - when this went, so did any pool of analysts and developers and training for them. I have no idea if this was ever replaced, but given that the government still hires (either directly, or via contracted IT companies), Business Analysts from outside, for internal development projects, it appears that they still don't train their own staff anymore. So they continue to rehire the same contractors who have been in every other project, again and again and again, turning it in to a career for these lucky people.
No one does nothing about it, because no one has any incentive to do so. I could suggest the reforms, which are obvious if you accept my analysis (and for obvious reasons, some people won't) .... but even if you do, there is no one willing to undertake them.
 |
| NHS APP - All Change - No Surprise. |
... today I wake up to the news that the UK has decided to scrap its government built, National Health Service (NHS) coronavirus mobile phone track and trace application.
Quelle Surprise ..... this was as certain an end decision, as the news that the sun comes up in the morning. The mobile app had been trialled on the Isle of White in the UK, with over 60,000 downloaders. But guess what, the app was reportedly plagued with technical faults, as well as user privacy concerns.
All this was predictable from the very start:
1). The privacy concerns were apparent before work even started, when the design model of using and storing the data in a centralised server was made known. The idea was that contact-matches were carried out on a remote server. Whilst this had some advantages (particularly in the amount of data epidemiologists would have available), this was also its weakness, as hacking concerns about that data were raised.
2). Due to Apples 'privacy' and code sharing restrictions (as the US government has found, Apple refuse to share development code or password cracking techniques), the NHS app was poor at recognising Apple's iPhones. So while the NHS app recognised about 75% of nearby Android handsets it only spotted around 4% of iPhones. Reportedly the third party code Bluetooth restrictions by Apple, led to this identification failures.
3). The UK's governments incredibly poor record at developing and introducing any software (more on this in a moment).
 |
| BBC Explanation Of Tracking App Approaches .... |
Why does this matter? Well track and trace is helpful requirement to help prevent a second wave of the coronavirus as lockdown eases. The idea is that mobile apps would log whenever two people have been in close proximity to each other for a certain period of time, so if they are later confirmed to have the coronavirus, anyone who had been in close contact would be advised that they should also get tested and/or self-isolate.
So what are we going to use now? Why the Google and Apple developed apps of course (maybe). These use the decentralised data response, where the contact-matching process is carried out on the handsets themselves, which although not collecting as much data, does respect privacy and possible hacking concerns about data being open to either Governments, and or hackers, to be de-anonymised and misused later.
Now, with this decision, its possible that not even this Google and Apple based app will be used in the UK .... Baroness Harding, who is heading the government response has said that "What we've done in really rigorously testing both our own Covid-19 app and the Google-Apple version is demonstrate that none of them are working sufficiently well enough to be actually reliable to determine whether any of us should self-isolate for two weeks [and] that's true across the world." ... so we wait and see.
In fairness it should be noted that other centralised apps, notably France (the other big promoter of that approach), have also had problems. The European Commission has said that France would face challenges over this approach, including incompatibility with other European countries apps, which are to be based upon the Google-Apple Exposure Notification API, on which they are building their own apps to limit the spread of Covid-19.
 |
| The UK Government Has An Epic List Of I.T. Project Failures ......... |
Finally, the reason why this expensive U-Turn is no surprise, is the UK governments historic and ongoing spectacular inability to develop and deliver IT projects on time and on budget.
The reasons for this are fairly well known and indeed I posted on it a decade ago, but never acted upon:
1) Lack of Civil Service Skills: The UK civil service hires its Business Analysts and other IT skills from outside as highly overpaid contractors. For reasons that are down to the way it organises itself, its never able to find or train specialists in its own systems, who can model how things currently work, and then detail how any changes would work.
I personally know of a former Cobol developer, who has worked solely in the UK governments employ as an IT contractor for over 20 years without break, at a mind boggling amount of money (we are talking at today's values of well over £1.25 million pounds of tax payers money). He's currently a highly paid 'Business Analyst' contractor (he just took the title instead of developer, and as he knows everyone, was simply re-employed on another government project).
The reasons for this Civil Service failure to produce its own Business Analysts or other IT specialists, is that those who regularly use and become experts in government systems such the benefit delivery or immigration software, work in local offices e.g. Department of Works and Pensions (DWP) ... whereas projects are devised and run by regional office staff who don't regularly use or understand those systems, and who are not impacted by any issues users uncover. See the Universal Credit project.
2) Ministerial Interference: What I.T. requirements that are initially submitted for development, are often changed on a political whim, to meet the ever changing political requirements of government ministers. This causes expensive and dangerous code rewrites ... after a certain point, the code has to be fudged, rather than re-written, because its too expensive to scrap it and start again. So its bad code, and a poor result. See the Universal Credit project.
3) The Project Managers and or IT companies that are contracted to run a project, all rehire the same developers, and analysts from the same small pool of 'good ol'boys,' who have been used on all the other projects. It doesn't matter that many of these have failed on other projects, or frankly couldn't make it in private industry, they just swirl around in the dirty waters of failed or poorly executed government projects over and over again.
4) No firm ownership or decision maker. Heads never roll for failures, as everyone points up the ladder, to a government minister who has usually long moved on. So no one ever feels the career consequences of failure. Consequently no one ever calls a halt to projects that have gone off track. So for example, it must have been apparent some months ago, that with many governments switching to the Google-Apple model (including Germany, Italy and Denmark), for the same reasons that we are now doing, it was likely that we would also be forced to do so. But we didn't stop until millions more pounds of development money had gone down the drain.
... so guess what? The UK Government IT projects all fail, or are poorly delivered, for all the same reasons, involving nearly all the same people and companies every time.
To give you an idea of the problems, the Institute for Government, recently reported that 80% of major government projects are at 'risk of failure' with most being classed as being either in doubt, hindered by problems, or virtually unachievable. This was excluding Brexit related projects, and was based upon the Infrastructure and Projects Authority's annual reports and referred to £455 billion of projects.
So why, given the literally billions of pounds of tax payers money that has been lost to these failures over the last 30 years, does no one reform the system?
Well, again, its inherent in the system that nothing will be ever done:
1) Government Ministers don't want their political hands tied by constraints such as cost or impacts of their late requirement changes.
2) Thousands of Civil Service staff are employed in various Government Projects. its a nice career location, with no public contact, and not strenuous. With no career damage for any failures, so why change? As its almost impossible to sack Civil Servants there is no fear of failure.
3) The governments own Civil Service IT specialists used to be in an organisation called ITSA - when this went, so did any pool of analysts and developers and training for them. I have no idea if this was ever replaced, but given that the government still hires (either directly, or via contracted IT companies), Business Analysts from outside, for internal development projects, it appears that they still don't train their own staff anymore. So they continue to rehire the same contractors who have been in every other project, again and again and again, turning it in to a career for these lucky people.
No one does nothing about it, because no one has any incentive to do so. I could suggest the reforms, which are obvious if you accept my analysis (and for obvious reasons, some people won't) .... but even if you do, there is no one willing to undertake them.
No comments:
Post a Comment
All comments are welcomed, or even just thanks if you enjoyed the post. But please make any comment relevant to the post it appears under. Off topic comments will be blocked or removed.
Moderation is on for older posts to stop spamming and comments that are off topic or inappropriate from being posted .... comments are reviewed within 48 hours. I don't block normal comments that are on topic and not inappropriate. Vexatious comments that may cause upset to other commentators, or that are attempting to espouse a particular wider political view, are reviewed before acceptance. But a certain amount of debate around a post topic is accepted, as long as it remains generally on topic and is not an attempt to become sounding board for some other cause.
Final decision on all comments is held by the blog author and is final.
Comments are always monitored for bad or abusive language, and or illegal statements i.e. overtly racist or sexist content. Spam is not tolerated and is removed.
Commentaires ne sont surveillés que pour le mauvais ou abusif langue ou déclarations illégales ie contenu ouvertement raciste ou sexiste. Spam ne est pas toléré et est éliminé.