Radio Frequency Interference in Breathalyzers
In February 1982, Smith & Wesson, a major manufacturer of DUI breath testing devices, notified law enforcement agencies that its Breathalyzer Model 1000 had been found to experience interference from radio frequency transmissions under certain DUI test conditions, resulting in false test results. Subsequent investigation suggested that the Breathalyzer Models 900 and 900A also were affected by various power levels.
The potential sources of radio frequency interference (RFI) within the law enforcement environment are numerous - for example, AM and FM radios, police station dispatchers, hand-held police transmitters, teletypes, and police radar units. Each of these may emit the kind of interference that could cause specific DUI breath alcohol test devices to render false results. It was then learned that the federal government (National Bureau of Standards) was testing for RFI sensitivity every type of breath testing device currently in use by law enforcement agencies in DUI investigations.
On September 10, 1982, Smith & Wesson issued a customer advisory suggesting that all breathalyzers used in drunk driving investigations be tested as a matter of preparedness for possible courtroom testimony to avoid the issue of RFI in breathalyzers.
Late in 1984, Smith & Wesson announced the availability of a kit designed to improve the Model 900A's resistance to RFI in DUi settings. In a customer advisory letter from Smith & Wesson's production manager, users of the machine were told that a "Kit I," consisting of a combination of a filter box and amplifier board was available for $147, and a "Kit II," consisting of only the filter box, was priced at $72. Additionally, the Model 900A could be converted by the factory for $171.
At about the same time, CMI came out with its state-of-the-art DUI breath testing instrument, the Intoxilyzer 5000. This new machine has an "RFI Detector." Obviously, the marketing of a kit and an "RFI Detector" to inhibit radio frequency interference constitutes a clear recognition of the problem. In view of such steps by manufacturers, it is difficult to imagine any prosecutor or witness in a DUi trial denying that the problem exists.
In May 1983, Don Nichols and Jim Kaster, Minneapolis DUI attorneys and editors of the Drinking/Driving Law Letter, obtained a copy of the National Bureau of Standards (NBS) preliminary report of test results for electromagnetic interference (EMI). The report was dated March 25, 1983. NBS was under contract with the National Highway Traffic Safety Administration (NHTSA) to conduct testing on evidential breath testers (EBTs) currently in use by law enforcement officials nationwide.
The report contained the "bottom line" results of RFI testing. Each of 16 unidentified instruments (the contract for testing specifically required that individual machines be identified by secret code only) were subjected to four different frequencies typically present in the police DUI investigation environment. Of the 16 units tested, only six showed minimal interference at each frequency; 10 of the 16 showed substantial susceptibility on at least one frequency. The report characterizes the potential effect of EMI on the testing of alcohol vapors in DUI cases as "severe." This characterization is found on page two of the report, where the writer is describing the first in-house demonstration of EMI to NHTSA representatives in March 1982:
"Early in 1982, the Washington D.C. Metropolitan Police Department reported to NHTSA that EBTs were found to display erroneous BAC [blood-alcohol content] readings in the presence of electromagnetic fields from radio transmission. NHTSA contacted LESL [Law Enforcement Standards Laboratory] and TSC [Transportation Systems Center], and it was agreed that if EBTs were susceptible to electromagnetic interference, action must be taken to solve the problem, and that it might be necessary to modify the NHTSA standard to include electromagnetic interference (EMI) susceptibility requirements. On March 24, 1982, representatives of NHTSA, TSC, and NBS were given a demonstration by police officers who routinely conduct breath testing using an EBT in a mobile van. One police officer operated his handheld radio within 0.3 m (1 ft.) of the EBT and demonstrated that the electromagnetic field could severely affect the analysis of alcohol samples."
As a part of Nichols and Kaster's FOLA, action, release was also sought of any preliminary drafts of the National Bureau of Standards (NBS) report on RFI and EMI, specifically, a preliminary report authored by Dr. John Adams of Boulder, Colorado. This report was rumored to be significantly different from the final report eventually released by the NHTSA. So different was the final edited version that it was reported that Dr. Adams asked to have his name removed from the list of those responsible, and, in fact, a review of contributors does not show the name of Dr. John Adams. The following are excerpts from the Adams report, which is entitled Effects for the Electromagnetic Fields on Evidential Breath Testers, written for the Electromagnetic Fields Division, National Bureau of Standards, Boulder, CO 80303:
These results cannot be extrapolated to predict how individual EBTs would be affected at other levels of field strength or at other frequencies.
These results show that EMI is a potential problem with many of the EBT units in current use. The additional measurement techniques, standards, and test levels suggested in the section on "Future Efforts Needed" are needed to fully evaluate the effects of EMI on EBTs.Unfortunately, there are a couple of problems with this crude procedure. First, only the frequency used by the testing device is being tested; the procedure does not determine if frequencies used by any of dozens of other EMF-emitting devices in the area will be detected. Second, the "calibration" is usually done with the breath machine turned on but not during an actual capture and analysis - i.e., during operating conditions in an actual DUI investigation.
The states may have to take interim measures to determine the extent of their individual problems with EMI affecting EBTs.
Dr. Richard F. Jensen, of Forensic Associates in Minneapolis, has offered the following comments concerning RFI:
Instrument failure because of radio frequency interference is a serious problem today. These problems can affect not only breath testing devices such as the Breathalyzer but also instruments in the forensic laboratories and clinical laboratories.
Radio frequency interference (RFI) is not a new or unique phenomenon. Communication systems employed by the armed forces in World War I experienced RFI, especially in areas where there was widespread use of radios in automobiles and airplanes. The complexity of radio communication has increased since that time, and the incidences of RFI have followed a similar pattern. A special-interest area of science, referred to as electromagnetic compatibility (EMC), has evolved out of this concern with RFI. In an effort to encompass the concerns of the entire electromagnetic spectrum, not just radio frequency, a broader designation has evolved: electromagnetic interference (EMI). (EMI and RFI will be used interchangeably in this discussion.) This discussion will deal specifically with radio frequency interference concerns: in evidential breath testing, in other forensic measurements, and in clinical laboratory measurements.
The preliminary research reported by the Electromagnetic Compatibility facility of the National Bureau of Standards has demonstrated that many evidential breath testing devices in DUI investigations experience EMI. At least 10 of the 16 instruments tested demonstrated some susceptibility to RFI. What this means, of course, is that the problem is one of a generic nature and is not limited to any specific instrument or manufacturer.
There are a number of opinions as to how radio frequency interferes with the performance of evidential breath test devices. The diversity of opinion results from the application of various theories and because the methods of measurement from one manufacturer to another differ in technology. In order to accurately evaluate the potential effect of radio frequency on evidential breath testing, it is imperative that each device be examined in its permanent location. A variety of procedures have been devised to accomplish this evaluation. All of these procedures are designed with a holistic approach where one must consider not only the instrument being tested but also its immediate surroundings and the involvement of the operator. The results of these site evaluations provide for a variety of important conclusions.
At the present time it may be that the breath testing instrument is the best potential detector for radio frequency in its environment. In keeping with a systems approach, the instrument will provide the precise data required for a site evaluation, particularly when coupled with a well designed evaluation protocol. All instruments are not equally susceptible to EMI. Therefore, it is appropriate to survey any given site with the instrument normally employed in that location.
Even if operators are well-trained and measurements are well documented in DUI investigations, RFI can occur. What then? There is absolutely no course of action except to decertify the affected instrument and immediately take it out of service. The instrument must remain out of service until the source of the interference is identified and eliminated or until the instrument can be modified to function in its environment. This course of action is dictated not only by the fact that these instruments are performing evidential analyses but also by the generally accepted practice of good analytical measurement. The concern should not be whether the results are high or low or whether they are to the benefit of the DUI defendant but simply whether or not they are accurate.
It is not my intent to discuss the details of the modification of breath test instruments in order to make them less susceptible to RFI. However, the task of modification is not an overly difficult one and should be undertaken if there is any indication that the instrument may potentially be affected by radio frequency in a manner that would cause erroneous results in DUI investigations. It is important to remember that the responsibility of electronic design is in the hands of the manufacturer and any subsequent modification must be coordinated by the manufacturer. Unfortunately, this is not often the case, Smith and Wesson has just recently offered a modification retrofit kit for the Model 900A.
The evidential testing of breath for alcohol concentration is not the only chemical test that may have the potential for interference of electromagnetic radiation. There are two additional areas that should cause intense concern. These are forensic analyses or examinations for the judicial system and clinical analyses for diagnostic purposes. The potential for interference is not only present in settings where these types of measurements are made, but there is the probability that the presence of an interference will cause a much larger deviation than those supposedly found in breath alcohol measurements. The Breathalyzer Model 900A, although capable of making accurate measurements, is not a sophisticated electronic instrument. Instruments such as gas chromatographs, mass spectrometers, and spectrophotometers are electronically more complex and therefore potentially more susceptible to the influence of electromagnetic radiation. In this great technological age there are literally an infinite number of instrumental measurements occurring on a regular basis. To attempt to discuss all of these measurements in detail would be impossible, but it is sufficient to state that there is the theoretical potential for radio frequency interference with each and every measurement. However, what should be of imminent concern are two specific areas of scientific measurement that could potentially affect each and every one of us directly at one time or another. As mentioned earlier, one area is forensic science and the other clinical chemistry. In the past decade both disciplines have enjoyed the benefits of great strides in technology. This is especially true in the complexity and number of instrumental measurements that are now being performed.Unfortunately, there are a couple of problems with this crude procedure. First, only the frequency used by the testing device is being tested; the procedure does not determine if frequencies used by any of dozens of other EMF-emitting devices in the area will be detected. Second, the
calibration is usually done with the breath machine turned on but not during an actual capture and analysis - i.e., during operating conditions in an actual DUI investigation.
The responsibility of forensic science is to provide the judicial system with the results of the scientific examination of physical evidence. These results are most often interpretations based on an instrumental analytical measurement. In addition to the classical types of examinations accomplished by the use of both the stereo and compound microscopes, the instruments employed by forensic scientists have expanded to a wide variety of sophisticated electronic devices of which many are computer controlled. These instruments are most often in laboratories that are associated with law enforcement personnel. With the increased use of communications equipment by law enforcement agencies in DUI cases and the proliferation of radio frequencies employed by these agencies, scientific equipment in the forensic science laboratory is being exposed to more and varying electromagnetic radiation than ever before. Because of this potential for radio frequency interference, the analytical measurements and the interpretations based on those measurements may be incorrect. Does this mean that under these circumstances there is no chance for an accurate measurement? Absolutely not. All that needs be done is to employ the rules that govern good analytical measurements in any area of science. First, all measurements must be performed at least in duplicate, meaning two separate measurements are made on the same sample. The purpose of this procedure is to substantially reduce the effect of a systematic error in DUi testing, of which RFI is but one example. Systematic errors are those that can arise from identifiable sources that can occur in a sporadic manner and that generally are the cause of large, unpredictable deviations of the measured quantity. Second, these duplicate analyses must be accompanied by well designed and systematic programs of quality control and proficiency testing.
The concerns outlined above can and should be expressed also for the clinical chemist and the measurements that are made on physiological samples for the purposes of diagnosing medical problems. The majority of clinical laboratories have sound programs of quality control and proficiency testing. However, there are very few clinical tests or procedures performed in duplicate, a practice that does not count for or ensure the absence of systematic error. Systematic errors can be from an infinite number of sources. The only way in which the presence or absence of a systematic error can be assured in an analytical measurement is by making that measurement in duplicate: two separate measurements on the same sample. Recently a sheriffs deputy in southern Minnesota related to me,
They don't like us to use our radios in the hospital; it louses up their tests. Need more be said?
Finally, let it be said that this concern for radio frequency interference in drunk driving investigations is not an overreaction. It is a scientific fact: One area of technology simply has not caught up with another. Radio frequency interference is a product of today's technological demands interacting with yesterday's design and manufacturing methods.
Note: The prosecution may counter the DUI defense attorney attack in a DUI trial by pointing out that the machine has an
RFI detector. The problem with such
detectors is that they are simply not reliable. First, as repeated tests have demonstrated, there is a segment of the frequency band to which the detector is essentially blind. If there is a source of interference from a device emitting electromagnetic waves in this frequency range, it will not be detected.
Second, the detectors are rarely calibrated correctly. Most law enforcement agencies are unqualified to perform a complete maintenance and calibration. Commonly, the instrument will be sent back to the manufacturer for annual calibration of the various
fail-safe devices (RFI detector, mouth alcohol detector, interferent detector, ambient air flag, etc.). In the meantime, the machine is often
calibrated by police officers by simply using something like a hand-held police radio. The device may be activated three or four times from three or four different angles and distances from the breath machine; if the RFI detector is triggered, it is considered
calibrated for DUI testing purposes
Unfortunately, there are a couple of problems with this crude procedure. First, only the frequency used by the testing device is being tested; the procedure does not determine if frequencies used by any of dozens of other EMF-emitting devices in the area will be detected. Second, the
calibration is usually done with the breath machine turned on but not during an actual capture and analysis - i.e., during operating conditions in an actual DUI investigation.