Category Archives: Law Enforcement IQ Trainwrecks

Calculation errors casts doubt on TSA Backscatter safety

Leave a reply

It is reported in the past week on Wired.com and CNN that the TSA in the United States is to conduct extensive radiation safety tests on their recently introduced backscatter full body scanners (affectionately known as the “nudie scanner” in some quarters).

An internal review of the previous safety testing which had been done on the devices revealed a litany of

  • calculation errors,
  • missing data and
  • other discrepancies on paperwork

In short, Information Quality problems. A TSA spokesperson described the issues to CNN as being “record keeping errors”.

The errors affected approximately 25% of the scanners which are in operation, which Wired.com identifies as being from the same manufacturer, and included errors in the calculation of radiation exposure that occurs when passing through the machine. The calculations were out by a factor of 10.

Wired.com interviewed a TSA spokesperson and they provided the following information:

Rapiscan technicians in the field are required to test radiation levels 10 times in a row, and divide by 10 to produce an average radiation measurement. Often, the testers failed to divide results by 10.

For their part, the manufacturer is redesigning the form used by technicians conducting tests to avoid the error in the future. Also, it appears from documentation linked to from the Wired.com story that the manufacturer spotted the risk of calculation error in December 2010.

Here at IQTrainwrecks.com we are not nuclear scientists or physicists or medical doctors (at least not at the moment) so we can’t comment on whether the factor of 10 error in the calculations is a matter for any real health concern.

But the potential health impacts of radiation exposure are often a source of concern for people. Given the public disquiet in the US and elsewhere about the privacy implications and other issues surrounding this technology any errors which cast doubt on the veracity and trustworthiness of the technology, its governance and management, and the data on which decisions to use it are based will create headlines and headaches.

 

The Wrong Arm of the (f)Law

Leave a reply
Courtesy of Steve Tuck and Privacy International comes this great story from the UK of how a simple error, if left uncorrected, can result in significantly unwelcome outcomes. It is also a cautionary tale for those of us who might think that flagging a record as being “incorrect” or inaccurate might solve the problem… such flags are only as good as the policing that surrounds them.
Matthew Jillard lives on Repton Road in a suburb of Birmingham. In the past 18 months he has been raided over 40 times by the police. During Christmas week he was raided no fewer than 5 times, with some “visits” taking place at 3am and 5am, disturbing him, his family, his family’s guests, his neighbours, his neighbour’s guests….
According to Mr Jillard,
9 times out of 10 they are really apologetic.
Which suggests that 1 time out of 10 the visiting police might annoyed at Mr Jillard for living at the wrong address(??)
The root cause: The police are confusing Mr Jillard’s address with a house around the corner on Repton Grove.
(scroll the map to the right to find Repton Grove)
Clancy Wiggum from the Simpsons
Not a spokesman for West Midlands Police

View larger map
Complaints to the police force in question have been met with apologies and assurances that the police have had training on how important it is to get the address right for a search. Some officers have blamed their Sat Nav for leading them astray.
Given the cost to the police of mounting raids, getting it wrong 40 times will be putting a dent in their budget. Also, the costs to the police of putting right any damages done to Mr Jillard’s home due to the incorrect raids (which have included kicking in his door at 3am on Christmas Day) will also be mounting up.
The police have said that “measures” have been taken to prevent Mr Jillard’s home being raided, including putting a marker against his address on the police computer systems. None of these measures appear to have stopped the raids, which come at an average frequency of more than one a fortnight (40 raids in 18 months).
This Trainwreck highlights the impact of apparently simple errors in data:
  1. Mr Jillard’s home is being disturbed without cause on a frequent basis
  2. His neighbours must be increasingly suspicious of him, what with the police calling around more often than the milkman
  3. The police force is incurring costs and wasting man power with a continuing cycle of fruitless raids.
  4. The real target of the raids are now probably aware of the fact that the police are looking for them and will have moved their activities away from Repton Grove.

Police Untelligence

2 Replies

From The Register comes this wonderful example of the problems that can arise where data is used for unintended purposes, resulting in poor quality outcomes for all involved.

The NYPD have been regularly raiding the home of an elderly Brooklyn couple. They’ve been hit 50 times over the past 4 years, which might mark them out as leading crime kingpins but for the fact that their address has wound up included in police data used to test notification systems. The Reg tags this as “a glitch in one of the department’s computers”, but Information Quality trainwreck observers will immediately recognise that the problem isn’t with the technology but with the Information.

The trainwreck is compounded by two facts which emerge in the article:

  1. NYPD believed that they had removed the couple’s address from the system back in 2007, but it appears to have not been the case (or perhaps it was restored from a backup)
  2. The solution the NYPD have now implemented is to put a flag on the couple’s address advising officers NOT to respond to calls to that address.

The latter “solution” echoes many of the pitfalls information quality professionals encounter on a daily basis where a “quick fix” is put in to address a specific symptom which then triggers (as el Reg puts it) “the law of unintended consequences”.  To cut through implication and suggestion, let’s pose the question – what happens if there is an actual incident at this couple’s home which requires a police response?

What might the alternative approaches or solutions be to this?

(And are the NYPD in discussions with the Slovak Border police about the perils of using live data or live subjects for testing?)

Slovak Police accidentally cause Terror Alert in Dublin

2 Replies

The Irish and International media have been busy the past few days covering the story of the horrendously botched security test by Slovakian Border Police which resulted in 90 grams of high explosive RDX finding its way to Dublin from Bratislava in the backpack of an unsuspecting Slovakian electrician who was travelling back to Ireland after Christmas at home. This lead to a street in Dublin City Centre being closed this past Tuesday, with homes and businesses evacuated, while police and Army bomb experts raided the innocent electrician’s home to secure the explosives.

A full timeline for the story can be found here.

This is a tale which has a number of classic elements of an IQTrainwreck about it. Continue reading →

The terror of the Terrorist Watch list

Leave a reply

A source who wishes to remain anoynymous sent us this link to a story on Wired.com about the state of the US Government’s Terrorist watch list.

The many and varied problems with the watch list have been covered on this blog before.

However, the reason that this most recent story constitutes an IQTrainwreck is that it seems that, despite undertakings to improve quality, the exact opposite has actually happened given:

  • The growth in the number of entries on the list
  • The failures on the part of the FBI to properly maintain and update information in a timely manner.

According to the report 15% of active terrorism suspects under investigation were not added to the Watch list. 72% of people cleared in closed investigations were not removed.

The report from the US Inspector General said that they “believe that the FBI’s failure to consistently nominate subjects of international and domestic terrorism investigations to the terrorist watchlist could pose a risk to national security.”

That quote sums up why this is an IQTrainwreck.

Continue reading →

These are the IQ trainwrecks in your neighbourhood

Leave a reply

Stumbled upon this lovely pictorial IQTrainwreck today on Twitter. Thanks to Angela Hall (@sasbi) for taking the time to snap the shot and tweet it and for giving us permission to use it here. As Angela says on her Twitpic tweet:

Data quality issue in the neighborhood? How many street signs (with diff names) are needed? Hmmmm

Data quality issue in the neighborhood? How many street signs... on Twitpic In the words of Bob Dylan: “How many roads must a man walk down?”

Finally caught… the most dangerous driver in Ireland

3 Replies

Yesterday’s Irish Times carried a story of how, after some extensive detective work, the Irish Police finally tracked down the most dangerous driver in Ireland, a Polish gentleman by the name of Prawo Jazdy.

This individual was given hundreds of speeding tickets and parking tickets over the past few years, and by June 2007 he had over 50 separate entries in the Irish Police computer system. So slippery was this offender that he kept giving police a different address almost every time he was stopped and ticketed.

He was finally brought to book by a quick thinking officer in the Traffic Corps of the Irish Police who tracked him down using a structured Master Data source. Surely this is an example of Information and Intelligence lead policing at its best? Continue reading →

Duff Data dumps 1 million on the dole (social security)… in France.

1 Reply

The Register carries a story this week that clearly shows  the impact of poor quality information on people, particularly in this time of tightening economic conditions when getting a job is hard enough.

It appears that the French Government’s Police Vetting database is not very complete or accurate. According to the French Data Protection authorities (CNIL), this highlights the

serious issues over the provenance of data illustrate all too clearly what happens when the government starts to collect data on its citizens without putting adequate measures in place for updating and accuracy checking.

It would appear that there are errors in 83% of records, with a range of degrees of seriousness. The errors in the database arise as a result of “the simple mechanism of mis-recording actual verifiable data”. In other words, poorly designed processes,  poorly designed data creation/maintenance processes, poorly trained staff, and a lack of information quality control contribute to the errors.

But what  of the cost to the French economy? Well, every person who has been blacklisted in error by this system is potentially drawing social security payments. On top of that they are not paying taxes into the French economy.

If, in a month, they are drawing €100 in social insurance instead of paying €100 in taxes, the net loss to the French economy is €200 per month.  €200 x 1Million =€200 million per month, or €2,400,000,000 per year.

So, on the basis of a very rough guesstimate, the value to the French economy of fixing these errors is €2.4billion per year. Is that a business case?

Grevious misuse of statistics

1 Reply

The BBC today (2008-10-23) carries a news story about the under-reporting of serious crime due to the mis-classification of certain crimes. The story is examined in more detail in an op-ed piece “How the police missed the violence”..

The problem has come about when the intended victim manages to escape serious harm when attacked. The example given concerns a pub fight in which the assailant attempts to “glass” the victim but she suffers only minor injuries not the intended major damage to her face. The police then classify the crime not as “grievous bodily harm with intent to cause serious injury.” but as merely(?) “grievous bodily harm”.

14 forces have gone back through their statistics and reclassified GBH’s as the more serious crime where intent was clearly present but no great injury was suffered. This has resulted in a 22% jump in reported violent crime over the same period last year although it has not affected the total number of offences reported.

The IQ lesson here concerns the way in which business terms are defined and then applied. The key word in the definition is “intent” but many officers on the ground concentrate on the phrase “to cause serious injury”.Mind you as an ordinary man in the street I think I would classify GBH as a violent crime but officially in falls into the “Other personal crime (with Injury)” category which is not seemed a serious category.

We all know toddlers are terrors…

5 Replies

… but this story (again culled from The Register.co.uk) shows the lengths the US government goes to to stop these known terrors from getting on planes via the Terrorist Screening lists maintained by the Dept Homeland Security.

“USA Today tells the story of a Disney World-loving 6-year-old who shares a name with someone on the “additional screening” list. Little John Anderson hasn’t made it onto the cleared list because his mum finds the TRIP web site confusing.”

So Little Johnny Anderson can’t fly to Disney World without having to prove he isn’t a terrorist. Hard to do for a tired and pissed off six year old, as any parent will tell you. I can only imagine how they explain it to him:

Mommy Anderson: Now Johnny, the nice TSA man only wants to make sure you won’t try to fly a plane into the Magic Kingdom and kill Goofy

L’il Johnny: Whhaaaaaahhhhhh…. iwannagoseegoofynmickeynminniendonalducknowmommy…whahhhh (Johnny kicks TSA agent… join us next week for L’il Johnny goes to Cuba).

Processes are in place for people to get themselves taken of the list via a website… but Johnny’s mother finds the website confusing. Having looked at the site it doesn’t seem to have any clear process entry point for “The US Government thinks (correctly) my child is a terrorist but good God I can keep them under control, please let us go to Disney”. It would seem that Mommy Anderson might need to get L’il Johnny to fill out a form to allow her to make a complaint on his behalf. He’s six. He may not understand that words like pursuant, perjury or that if he lies on the form he could to prison until he is ELEVEN!!

So the process to correct errors in the information is not customer focussed. Mommy Anderson and L’il Johnny just have to put up with his being on a watch list because the process to correct the list isn’t friendly to the information consumer/creator.

L’il Johnny isn’t alone. The Register points out that he shares his story with Javaid Iqbal, a seven year old British boy who was stopped repeatedly at US airports when on holiday in Florida (which if you are seven years old translates as Disneyland). He shared his name with someone who had been deported from the US. His name was shared, not his age. L’il Javaid’s passport is now stamped that he underwent high level security checks, pretty much condemning him to a life of Eurostar trips to EuroDisney from now on. Imagine if his name had been Lee Harvey Oswald.

These are the easy funny stories that highlight weaknesses in the quality of information and processes in this important function of the US Dept of Homeland Security. A less funny story is that

  1. There is some confusion about how big these lists are… a government report says 7555000, but a spokesman for the TSA says that he thinks it is less than half that
  2. 97% of people who have managed to get through the process that Mommy Anderson finds confusing are simply namesakes of people on the lists.
  3. Less than half of the requests to be taken off the lists (to correct inaccuracy in information and improve quality) have been processed since the facility went ‘live’ in February.  41% are “still being discussed” or are awaiting further documentation to prove the claims of the people who complained.

Why is this an IQTrainwreck?

  1. The level of inaccuracy in the information causes disruption to people… L’il Johnny or L’il Javaid won’t understand what is going on. And in Javaid’s case it may actually affect his ability to re-enter the US (or other countries) in later life.
  2. There appears to be a lack of consolidated governance and control… not being able to answer “how many names are on the lists” and get a consistent response is like asking a company how many customers are in their CRM systems and getting mixed responses….( ohhh – bad example).
  3. The processes for correcting information are not ‘customer friendly’ and don’t seem to cater for the existence of children with names who might be too young to understand the processes or even the forms. The fact that parents find the process difficult to navigate suggests there is scope for improvement.

In any other context I’d be in favour of any measure that keeps screaming children off planes, particularly long flights (by which I mean any flight that lasts longer than 6 minutes).

However, I am reminded in these cases of why my father (a civil servant in the Irish Republic and, at the time, a senior trade unionist) used to hate going on trips to the North of Ireland.  Every so often some British Army squaddie or RUC officer would take him out of the car or off the bus for questioning because:

  • He had glasses
  • He had a beard
  • He wasn’t speaking with an English accent
  • therefore he must be Gerry Adams, leader of Sinn Fein and his id papers must be faked.

My parents had a hard time explaining that to the snotty-nosed crying kids in the back seat…