Previous in Forum: Could This Planet Physically Exist?   Next in Forum: Fastest Bicycle On Earth
Close
Close
Close
12 comments
Rate Comments: Nested
Associate

Join Date: Jul 2013
Posts: 33

Statistical Sampling in Security

11/08/2014 10:46 PM

We were taught about concept of SQC (Statitistical Quality control) where sampling is used instead of every piece for quality assurance, with reasonable accuracy. Can we use a similar concept in a different environment. Delhi metro carries about 2.7 million passengers every day. Each one of these passengers is passed through a metal detector gate and individually frisked. Further each piece of baggage (about 2.0 million pieces every day, including women hand bags) is X-rayed on conveyor type machines. This creates lot of chaos. I do not know what is success rate of such security arrangement in catching risks, but the quality of frisking and X-ray is not maintained due to such large numbers. I wonder whether statistical sampling could be employed in such cases. I do not know whether those responsible for security can emulate example of "London Tube", which has no such checks.

Register to Reply
Interested in this topic? By joining CR4 you can "subscribe" to
this discussion and receive notification when new comments are added.

Good Answers:

These comments received enough positive votes to make them "good answers".

"Almost" Good Answers:

Check out these comments that don't yet have enough votes to be "official" good answers and, if you agree with them, vote them!
Guru

Join Date: Oct 2008
Posts: 42355
Good Answers: 1693
#1

Re: Statistical sampling in security

11/08/2014 11:39 PM

Use it if you are satisfied with "reasonable" security.

Register to Reply
Guru
India - Member - New Member

Join Date: Sep 2009
Location: Temporarily at Ashburn, VA
Posts: 2744
Good Answers: 164
#2

Re: Statistical sampling in security

11/09/2014 1:46 AM

This happens figuratively all over the place (My apologies Mr. Saxena, just feeling mighty happy this Sunday)..

Security

__________________
Nothing worthwhile can ever be taught, it can only be learnt.
Register to Reply
Guru

Join Date: Sep 2012
Location: Iowa, USA
Posts: 577
Good Answers: 50
#3

Re: Statistical sampling in security

11/09/2014 6:03 AM

You have to perform security per your countries regulations. You can't invent your own security procedures. The end destination country can place requirements on departure flights.

In the USA, the rules flow down from the Federal Code of Regulations, which puts requirements on the Transportation Security Administration (TSA). The TSA then has procedures that they can modify e.g. what you can carry on. Every carry on/checked bag gets X-rayed. All persons go through metal detectors.

So since your asking, you must be in a country that is giving you license to make your own security rules. Where is this?

A bomb on ground transportation has limited catastrophic outcome vs a loaded aircraft. You can not use that as any rational reason for aircraft security.

Follow the regulatory requirements of your country.

__________________
ignator -
Register to Reply Score 1 for Off Topic
Guru
United Kingdom - Member - Not a New Member Hobbies - Musician - New Member Hobbies - Fishing - New Member

Join Date: May 2006
Location: Reading, Berkshire, UK. Going under cover.
Posts: 9684
Good Answers: 468
#4
In reply to #3

Re: Statistical sampling in security

11/09/2014 6:42 AM

OP is talking about Metro travel. Underground trains. Subway.

__________________
"Love justice, you who rule the world" - Dante Alighieri
Register to Reply
Guru

Join Date: Sep 2012
Location: Iowa, USA
Posts: 577
Good Answers: 50
#6
In reply to #4

Re: Statistical sampling in security

11/09/2014 10:37 AM

I somehow missed that detail of "metro", and thought that he was using ground transport as rational for air travel.

What really sucks is that we used to live in a free society where terrorism was not a worry.

__________________
ignator -
Register to Reply
Guru

Join Date: Mar 2012
Location: Out of your mind! Not in sight!
Posts: 4424
Good Answers: 108
#9
In reply to #6

Re: Statistical sampling in security

11/09/2014 8:30 PM

Yes right after a world war and a cold war. You maybe have been growing up in a quiet part of the world for a couple of years and now worlds end is catching up with you. You can call this century the Terror War century

__________________
Common Sense Dictates
Register to Reply
Guru

Join Date: Feb 2011
Posts: 1119
Good Answers: 11
#5

Re: Statistical sampling in security

11/09/2014 9:31 AM

Quantitative and qualitative approach are two different thing and cannot be used in the same sense.

__________________
" To infinity and beyond" - Buzz Lightyear
Register to Reply
3
Guru
Engineering Fields - Power Engineering - New Member

Join Date: May 2007
Location: NYC metropolitan area.
Posts: 3230
Good Answers: 444
#7

Re: Statistical sampling in security

11/09/2014 12:09 PM

No, One aspect of SQC relies on uniform samples in uniform batches, hardly the way people carry goods with them. Second, if you blindly sample 10% of the population then your expectation is that you will only detect 10% of the goal items, while letting 90% get by your screening.

Your adversaries will always use the complement of your screening process; i.e., if they determine that you are screening to 99% then they will try to pass 1,000 devices on the expectation that 10 will avoid detection.

__________________
“Tell me and I forget. Teach me and I remember. Involve me and I learn.” Ben Franklin.
Register to Reply Good Answer (Score 3)
Guru
Hobbies - DIY Welding - Don't Know What Made The Old Title Attractive... Popular Science - Weaponology - New Member United States - US - Statue of Liberty - 60 Year Member

Join Date: Apr 2009
Location: Yellowstone Valley, in Big Sky Country
Posts: 7425
Good Answers: 295
#12
In reply to #7

Re: Statistical sampling in security

11/10/2014 1:03 PM

"Your adversaries will always use the complement of your screening process..."

Absolutely spot on correct. Any weakness that can be identified will be exploited by anyone determined to defeat the system.

__________________
Semper Ubi Sub Ubi
Register to Reply
Guru
United Kingdom - Member - Not a New Member Hobbies - Musician - New Member Hobbies - Fishing - New Member

Join Date: May 2006
Location: Reading, Berkshire, UK. Going under cover.
Posts: 9684
Good Answers: 468
#8

Re: Statistical Sampling in Security

11/09/2014 6:00 PM

If a CD or DVD doesn't play, or a breadmaker doesn't make bread, the customer mutters, then returns the purchesed item and receives a replacement or a refund. If a machine-screw is badly threaded, and won't fit corrrectly, the operative mutters, then bins it and picks up another.

If a suicide bomber slips through the sampling, there will probably be any people killed.

Not quite on the same scale.

__________________
"Love justice, you who rule the world" - Dante Alighieri
Register to Reply
Guru
Safety - Hazmat - New Member United States - US - Statue of Liberty - New Member Engineering Fields - Chemical Engineering - Old Hand

Join Date: Mar 2011
Location: Lubbock, Texas
Posts: 14331
Good Answers: 162
#10

Re: Statistical Sampling in Security

11/10/2014 11:50 AM

I think it unwise to do anything less than 100% screening for reasons mentioned by other posters. I also think the way the Israelis undertake this is really very clever. Talk to each passenger, gauge reactions, understand body language, and also use a "sniffer" to detect the smell of fear. If a dog can detect fear, humans eventually will get it. General profiling as to the type of character this usually represents may be helpful in funneling the "applicants" to the correct "reception" area.

Also, let the bad guys know that if they that if selected for the advanced probing, they will receive a pig lard suppository right before the chamber detonates their bomb vest.

__________________
If it ain't broke, don't fix it. Just build a better one.
Register to Reply Score 1 for Good Answer
2
Guru
Hobbies - Musician - New Member

Join Date: Feb 2012
Posts: 523
Good Answers: 17
#11

Re: Statistical Sampling in Security

11/10/2014 12:19 PM

GAs to #7 from RAMConcule and #10 James Stewart:

Risk is a major consideration. If one Type II error occurs, you say this is a good guy when he is really a bad guy, can you tolerate the cost of this mistake?

SPC is used to demonstrate that processes in a state of statistical control have not drifted from that state. There is a lot of power in using SPC, but SPC methods will not control your process any more than measuring your blood pressure will control your blood pressure. You place controls on your process, you place controls on your blood pressure, and use the tools to see if the controls are working.

I am not sure you could demonstrate you have a contolled process of delivering "good" passengers into the metro system. Even if you could, SPC doesn't prevent Out of Control conditions. SPC helps you understand if your out of control condition is likely from a special cause or a common cause. What is the cost of even one statistical outlier? The value you put on that tells you how much to spend on your controls.

Register to Reply Good Answer (Score 2)
Register to Reply 12 comments

Good Answers:

These comments received enough positive votes to make them "good answers".

"Almost" Good Answers:

Check out these comments that don't yet have enough votes to be "official" good answers and, if you agree with them, vote them!
Copy to Clipboard

Users who posted comments:

Doorman (1); IdeaSmith (1); ignator (2); James Stewart (1); JohnDG (2); kvsridhar (1); lyn (1); Noudge79 (1); RAMConsult (1); reward54 (1)

Previous in Forum: Could This Planet Physically Exist?   Next in Forum: Fastest Bicycle On Earth

Advertisement