SCRAM: An Investigation
PIRE is currently conducting studies on the reliability of available transdermal alcohol sensors. Apparently the results are just examples not conclusions until the ever so trustworthy NHTSA reviews them. Their findings thus far include some theories that I had not yet encountered, such as the original tests claim :
Case Study 1. Elevated TAC Reading
On 12/30/2002, the SCRAM Network generated an Alcohol Detected alert for Client 1. This
alert is severe in nature and should be analyzed immediately.
An Alcohol Detected alert is generated when the SCRAM bracelet detects ethanol over the
typical agency standard of .02%. An interferant can also cause elevated TAC readings. By
analyzing the alert, a real drinking event can quickly be differentiated from an alert that is
caused by use of a banned substance or an environmental factor.
PIRE has found that Detection algorithm is good not perfect – it can
misclassify rapid rise in BAC as an external interferent.
An example of a false-negative, however this device is spot-on accurate remember! A good look at A
MS's perfume test shows a fast paced decline in which they claim differen- tiated the interferent from a drinking episode, but pulling the bracelet up tight could retard diffusion and create the curve. Had they not known the environmental factors I am confident they would have prosecuted on the curve. This is only one example, there are thousands of tests per day in a non-controlled environment.
Here are some of PIRE's findings thus far:
• Minor problems included:
– SCRAM: some bouncing on ankle while exercising
– SCRAM: minor delays at airport security (5-7 min)
– SCRAM: minor bruising two female subjects
– WrisTAS: raising rash on skin surface
– WrisTAS: pads get stinky after a week
– WrisTAS: must remove for showering
Some SCRAM-related issues:
• Data may get spiky at times – probably water
accumulating in the sensor – accuracy suffers
• If water is present, sensor may lose ability to detect
ethanol or have delayed sensitivity – with a recent
design change this may now be better
• Paced drinking with food may not trigger an "alert"
• Detection algorithm is good not perfect – it can
misclassify rapid rise in BAC as an external interferent –
their algorithm tries to protect against false positives
• Modem communications usually work well, but they do
not work with mobile phones – this will be problem if no
landline available to offender.
• A few landlines had trouble dialing out.
Conditions of the skin also distort readings, and the accumulation of water in the bracelet ( if the reads are spikey, unknown environmental factors may play a role-RFI) brings me back to my original hypothesis of erroneous obstruction reads. This aspect of the device works like this:
As the IR sensor shines an infrared beam against the client’s skin, the skin absorbs a certain
amount of the signal. The rest of the signal is reflected back to a receiver in the SCRAM
bracelet. The receiver measures the amount of signal that was reflected, and converts that
signal to a voltage. Initially, the voltage is used to establish the amount of signal that is
typically reflected off of that particular client’s skin. Subsequent signals are compared against
this initial signal. If the subject inserts a foreign object between the SCRAM bracelet and the
skin, the strength of the reflected infrared signal changes. If the subject removes the SCRAM
bracelet, the SCRAM bracelet detects the absence of an infrared signal, and generates a
Potential Removal alert.
Hypothesis: Dehydration, dry skin, or water trapped in the bracelet (sensor) may alter the amount of the signal absorbed by the skin, thus therefore resulting in an erroneous read. Also, it may only take a small portion of the sock under the corner of the bracelet to move the sensor and register as obstruction, with no way to prove this was the case they would believe it was intentional.
AMS reports the IR distance differential for different types of obstructants in volts:
Client 1) Normal = 3.58 - 4.11
Black trash bag = 3.05 - 3.12
difference -.53 - .99 (assuming the guage is difficult,these are examples of the sensitivity)
Client 2) Normal = 4.16 - 4.29
Saran wrap = 4.29 - 4.87
difference +.13 - .56 volts
Client 3) Normal = 3.55 - 4.02
Thick Sock = 5.00 - 5.00
difference +1.45 - .98 volts
These experiments were conducted in alliance with the consumption of alcohol, and in all of these cases, although the TAC/BAC correlation was distorted, there was an indication of consumption. Unfortunately the sock was the most effective interference, the common everyday apparel. Client 3 shows a thick sock has a dramatic effect on the read, a guage which exceeded this devices measuring ability, leaving me with the conclusion that the top of the sock so much as in the corner of the bracelet can produce an erroneous read, and pant legs, blankets, etc...
With 2,200 people wearing SCRAM tested twice an hour, that is 105,600 tests daily so the importance of reliability is overwhelming. If SCRAM were 99.9% effective this would mean that today there were over 100 false positive/negative reads. Their claim is that it tests for alcohol twice an hour, so this may not include the obstruction tests. As a result of the saran wrap test they quote," The abrupt and sustained peaking of the IR Distance Voltage readings proves that
the subject used an obstruction to mask the presence of alcohol." According to the sock test and my experience this proves nothing, it gives reasonable suspicion, probable cause and is a cause for action (not enough for conviction), making it necessary to find corroborating evidence such as urinalysis, PBT, even a paper trail. The alcohol could be from the spouse spraying disinfectant and the vulnerability of the exposed bracelet from a sock just convicted the innocent!
REFERENCE:
http://www.nlectc.org/training/commcorr2005/marques_comcorr2005_transdermal.pdf
http://j.b5z.net/i/u/2099113/i/TamperDetectionWhitePaper.pdf
Lasley, 20, signed up for the anklet because he hopes a Durham judge will give him a lighter sentence this month on a charge of drunken driving _ he faces a year in prison...The anklet is also sensitive to any alcohol that comes in contact with the skin. For example, Lasley's anklet registered when he put on a shirt washed in a detergent that contained alcohol. http://www.statesville.com/servlet/Satellite?pagename=SRL%2FMGArticle%2FWSJ_BasicArticle&c=MGArticle&cid=1031783999368&path=!statenews
HOME
Case Study 1. Elevated TAC Reading
On 12/30/2002, the SCRAM Network generated an Alcohol Detected alert for Client 1. This
alert is severe in nature and should be analyzed immediately.
An Alcohol Detected alert is generated when the SCRAM bracelet detects ethanol over the
typical agency standard of .02%. An interferant can also cause elevated TAC readings. By
analyzing the alert, a real drinking event can quickly be differentiated from an alert that is
caused by use of a banned substance or an environmental factor.
PIRE has found that Detection algorithm is good not perfect – it can
misclassify rapid rise in BAC as an external interferent.
An example of a false-negative, however this device is spot-on accurate remember! A good look at A
MS's perfume test shows a fast paced decline in which they claim differen- tiated the interferent from a drinking episode, but pulling the bracelet up tight could retard diffusion and create the curve. Had they not known the environmental factors I am confident they would have prosecuted on the curve. This is only one example, there are thousands of tests per day in a non-controlled environment.Here are some of PIRE's findings thus far:
• Minor problems included:
– SCRAM: some bouncing on ankle while exercising
– SCRAM: minor delays at airport security (5-7 min)
– SCRAM: minor bruising two female subjects
– WrisTAS: raising rash on skin surface
– WrisTAS: pads get stinky after a week
– WrisTAS: must remove for showering
Some SCRAM-related issues:
• Data may get spiky at times – probably water
accumulating in the sensor – accuracy suffers
• If water is present, sensor may lose ability to detect
ethanol or have delayed sensitivity – with a recent
design change this may now be better
• Paced drinking with food may not trigger an "alert"
• Detection algorithm is good not perfect – it can
misclassify rapid rise in BAC as an external interferent –
their algorithm tries to protect against false positives
• Modem communications usually work well, but they do
not work with mobile phones – this will be problem if no
landline available to offender.
• A few landlines had trouble dialing out.
Conditions of the skin also distort readings, and the accumulation of water in the bracelet ( if the reads are spikey, unknown environmental factors may play a role-RFI) brings me back to my original hypothesis of erroneous obstruction reads. This aspect of the device works like this:
As the IR sensor shines an infrared beam against the client’s skin, the skin absorbs a certain
amount of the signal. The rest of the signal is reflected back to a receiver in the SCRAM
bracelet. The receiver measures the amount of signal that was reflected, and converts that
signal to a voltage. Initially, the voltage is used to establish the amount of signal that is
typically reflected off of that particular client’s skin. Subsequent signals are compared against
this initial signal. If the subject inserts a foreign object between the SCRAM bracelet and the
skin, the strength of the reflected infrared signal changes. If the subject removes the SCRAM
bracelet, the SCRAM bracelet detects the absence of an infrared signal, and generates a
Potential Removal alert.
Hypothesis: Dehydration, dry skin, or water trapped in the bracelet (sensor) may alter the amount of the signal absorbed by the skin, thus therefore resulting in an erroneous read. Also, it may only take a small portion of the sock under the corner of the bracelet to move the sensor and register as obstruction, with no way to prove this was the case they would believe it was intentional.
AMS reports the IR distance differential for different types of obstructants in volts:
Client 1) Normal = 3.58 - 4.11
Black trash bag = 3.05 - 3.12
difference -.53 - .99 (assuming the guage is difficult,these are examples of the sensitivity)
Client 2) Normal = 4.16 - 4.29
Saran wrap = 4.29 - 4.87
difference +.13 - .56 volts
Client 3) Normal = 3.55 - 4.02
Thick Sock = 5.00 - 5.00
difference +1.45 - .98 volts
These experiments were conducted in alliance with the consumption of alcohol, and in all of these cases, although the TAC/BAC correlation was distorted, there was an indication of consumption. Unfortunately the sock was the most effective interference, the common everyday apparel. Client 3 shows a thick sock has a dramatic effect on the read, a guage which exceeded this devices measuring ability, leaving me with the conclusion that the top of the sock so much as in the corner of the bracelet can produce an erroneous read, and pant legs, blankets, etc...
With 2,200 people wearing SCRAM tested twice an hour, that is 105,600 tests daily so the importance of reliability is overwhelming. If SCRAM were 99.9% effective this would mean that today there were over 100 false positive/negative reads. Their claim is that it tests for alcohol twice an hour, so this may not include the obstruction tests. As a result of the saran wrap test they quote," The abrupt and sustained peaking of the IR Distance Voltage readings proves that
the subject used an obstruction to mask the presence of alcohol." According to the sock test and my experience this proves nothing, it gives reasonable suspicion, probable cause and is a cause for action (not enough for conviction), making it necessary to find corroborating evidence such as urinalysis, PBT, even a paper trail. The alcohol could be from the spouse spraying disinfectant and the vulnerability of the exposed bracelet from a sock just convicted the innocent!
REFERENCE:
http://www.nlectc.org/training/commcorr2005/marques_comcorr2005_transdermal.pdf
http://j.b5z.net/i/u/2099113/i/TamperDetectionWhitePaper.pdf
Lasley, 20, signed up for the anklet because he hopes a Durham judge will give him a lighter sentence this month on a charge of drunken driving _ he faces a year in prison...The anklet is also sensitive to any alcohol that comes in contact with the skin. For example, Lasley's anklet registered when he put on a shirt washed in a detergent that contained alcohol. http://www.statesville.com/servlet/Satellite?pagename=SRL%2FMGArticle%2FWSJ_BasicArticle&c=MGArticle&cid=1031783999368&path=!statenews
HOME
posted by marcellus91872 at 12:59 AM
8 Comments:
A shirt washed in detergent that contained alcohol? That puts my mind at ease. I am hours away from meeting with officials because of a "confirmed consumption" reading. This reading was in the middle of my work day, while I was with patients in a dental clinic. When officials tried to have me sign a form stating that I would "not posses, use, or come in contact with any substance containing alcohol" I laughed. Do they want me to spread infectious diseases to my patients? I refused and have been in contact with the official monitoring my case. She had readings throughout the duration of the scram's application, but none were "confirmed consumption". I am very angered because of these happenings, as I have been sober for 5 months.
hOW MUCH DOES A WRISTAS Cost ?
I just had my "bracelet" read and got some "numbers" registered. I had some balsamic vinegar on my salad at the time that it showed some activity and ate some chocolate at the other time that activity was registered. What the hell is up with that?? So then they made me pee and blow. Any input on this would be appreciated.
I have yet to see Wristas marketed, probably external influence would be harder to avoid considering the location it is worn.
As far as the reads, I don't know the numbers or burn-off so it is impossible to discuss possibilities. More than likely an external source, laundry detergent was the alledged contaminent in one read. I only eat balsamic vinegar and oil on my salad and did not register due to the consumption of any food, it is possible however, endogenous does happen.
Marcellus
My brother just got a call from his PO stating that the report she received, showed some kind of obstruction was made to the bracelet and that the report showed no activity between the hours of 9:15 p.m. to 5:27 a.m. in the morning. Has anyone ever had a NO reading? I mean how can one prove that they were sleeping at the time? Absolutely no drinking or intentions where made to obstruct the bracelet, yet my brother might go to prison.
Hey all,
I am on the SCRAM here in Michigan, I have not drank any alcohol since I was put on probation a yr ago. I have been wearing the SCRAM for 7 months, and in this time it has had false readings of alcohol consumption. The last 3 being in the past 2 weeks. I am at a loss because my PO is looking to violate me. One reading was .032. That was the first being in the middle of May. I have not drank any alcohol, yet it comes up with a positive. The past one was yesterday, last night and today. Reading alcohol up to .08. I even blew into the pbt and came up with .000, yet my PO says the SCRAM does not lie. I know POs don't give any credibility to the people on probation, however, I have found myself going to the police station and paying to blow for an alcohol test to give me something else to prove I am not drinking. I went Monday night and blew .000 but the scram was reading alcohol, but my PO would not tell me what the reading was. He seemed annoyed that I am going to the Police station and blowing and told me that the SCRAM will hold up in court against a PBT, is this true???? This morning I went to the POs office to have the SCRAM checked out, it showed alcohol all the way throughout the night. I had the PBT from the police station and I even had to blow at the POs office, again I blew .000, but the SCRAM was detecting alcohol. I am at a loss as to what to do. He did not change out the SCRAM, he just told me he would be in touch. WTF. I know this much about myself to say that if I am going to drink, I will be on a 4 day binge and the SCRAM will be reading 4 digits as well. I am not going to risk my freedom over a beer. I am charged with beating hell out of 2 guys, one a reserve cop, (just my luck). Prosecutor said I didn't fight fair, and here I am. I don't mess around with drugs, and if I did drink, I would say so. My word is worth more to me than being caught in a lie. WHAT RECOURSE DO I HAVE WHEN I BELIEVE THE SCRAM IS EITHER FAULTY, OR SOMEONE IS PLAYING WITH THE COMPUTER.
Scram only reads up to .08, after that it is a mystery, was it .999, noone knows! Get a lawyer who is familiar, get the officers to testify, during cross examination drill the P.O. if you seemed intoxicated, had the scent of alcohol etc... & hope your judge has more integrity than the P.O.
Keep getting evidence that you are not drinking (I never had that chance) and you may get a law suit out of it. He is taking PBT's, face to face, he'll get his!
Scram only reads up to .08, after that it is a mystery, was it .999, noone knows! Get a lawyer who is familiar, get the officers to testify, during cross examination drill the P.O. if you seemed intoxicated, had the scent of alcohol etc... & hope your judge has more integrity than the P.O.
Keep getting evidence that you are not drinking (I never had that chance) and you may get a law suit out of it. He is taking PBT's, face to face, he'll get his!
Post a Comment
<< Home