There
is an underlying premise the DUI defense practitioner should act on
that all of the states approved breath test machines are run by computer
software with source codes that translates the breath results to BAC,
and these source codes have human made errors. It is estimated
that for every 1000 lines of code there are approximately 25 errors.
[The Science of Debugging, Telles and Hseith, Coiolis Technology Press,
2001, page 57.] A modern breath machine like the Draeger 7110 has over
53,000 lines of code. Even attempting to correct errors in the
source code can be problematic. Studies have researched the new error
rate that occurs when an effort to fix the program is attempted,and
new problems arise 15-50% of the time.
A
source code problem has been identified with the “buffer overflow”
from theAlcotest 7110 MKIII-C breath
machine manufactured by Draeger Safety Diagnostics, Inc
The 7110 analyzes each breath
sample with two different methods: Infrared analysis and fuel cell analysis.
A buffer overflow error occurs when the infrared analysis and the fuel
cell analysis in the first two breath samples of the subject fall outside
the accepted the range. The subject must then give a third breath sample.
When this occurs there is a problem because the 7110 software was designed
to preserve only four readings. Buffer overflow error results with the
machine overlooking the second fuel cell analysis.
A
challenge to this testing equipment was made in the case of New Jersey
v. Chun (2008) 194 N. J. 54. The Chun
court declined to require a change in the 2100 to 1 blood to breath
ratio the 7110 uses to convert a breath result into a blood alcohol
level, finding that the ratio gives most defendants an advantage as
the 2100 to 1 ratio slightly understates the partition ratio of most
people, while acknowledging that a small number of people will be disadvantaged
by a device that uses the 2100 to 1 blood/breath ratio, Id. at
96. The Chun Court declined to mandate that the state put breath
temperature sensor devices on their breath machines to measure the temperature
of a subject’s breath. The court acknowledged that there is an argument
that for every 1° Celsius increase in a subject’s breath temperature,
a 6.8% increase in the breath alcohol analysis could occur. The Chun
Court felt that there was insufficient evidence in the record to show
that falsely high readings occurred because of increased breath temperature.
Id. at 109. The court’s reasoning was that the potential
for artificially high readings was alleviated by using the 2100 to 1
ratio; that the ratio already took temperature variance into consideration;
and that the practice of truncating the third digit of breath results
also protected the defendant. Id. at 106-108. Another imperative
finding of the Chun court was the discovery of an algorithm in
the 7110 software that compensated for an occurrence known as fuel cell
drift that as the cells age their reaction to alcohol slows down.
Id. at 121-124. The Chun Court ended up adopting the
findings and conclusions of the Court’s Special Master as modified.
Id. at 149.
The
DUI defense lawyer should exploit the fact that the 7110
that does not have a breath temperature sensor device installed on it.
Further, in any other state than New Jersey with have a 60+ female client
and charged with a refusal in a breath case, the Chun decision
to the court’s attention to strike the refusal.
|
