Sensor system for passive in-vehicle breath alcohol estimation

Abstract

To provide a method for quickly estimating an alcohol level in a driver's breath.SOLUTION: There is provided a method for passive breath alcohol detection comprising: A) passively obtaining a first air sample from an air inside a vehicle; B) determining levels of (i) a tracer gas (CoTG) and (ii) analyte (CoA) present in the first air sample; E) continuing to passively obtain an N number of air samples from the air inside the vehicle, and for each air sample obtained, determining the levels of CoTG and CoA present in the air sample; F) determining a number of level peaks in the CoTG and a number of level peaks in the CoA present in each of the air samples; G) determining a confidence interval on the basis of the number of level peaks in the CoTG and the number of level peaks in the CoA; and H) controlling operation of the vehicle on the basis of a function of the confidence interval and CoA present in the air samples.SELECTED DRAWING: None

Description

[Technical Field]

[0001] References to pending prior patent applications This patent application: (i) A continuation-in-part of pending prior U.S. patent application Ser. No. 17 / 462,318 (Patent Attorney Docket No. ACTS-108460-0009-101 CON), filed on August 31, 2021, for "SENSOR SYSTEM FOR PASSIVE IN-VEHICLE BREATH ALCOHOL ESTIMATION" by Automotive Coalition for Traffic Safety, Inc., which patent application: (a) A continuation of prior U.S. patent application Ser. No. 15 / 389,724 (Patent Attorney Docket No. ACTS-108460-0009-101), filed December 23, 2016, by Automotive Coalition for Traffic Safety, Inc., for "SENSOR SYSTEM FOR PASSIVE IN-VEHICLE BREATH ALCOHOL ESTIMATION," which patent application: (1) Claiming the benefit of prior U.S. Provisional Patent Application No. 62 / 312,476, filed March 24, 2016 (Patent Attorney Docket No. ACTS-108460-0009-101), for "SENSOR SYSTEM FOR PASSIVE IN-VEHICLE BREATH ALCOHOL ESTIMATION" by Automotive Coalition for Traffic Safety, Inc.; (ii) Claims benefit of pending prior U.S. Provisional Patent Application No. 63 / 562,889, filed March 8, 2024 (Patent Attorney Docket No. ACTS-1 PROV), for "SENSOR SYSTEM FOR PASSIVE IN-VEHICLE BREATH ALCOHOL ESTIMATION" by Automotive Coalition for Traffic Safety, Inc.

[0002] The four above-identified patent applications are incorporated herein by reference. The present invention relates to a method for detecting the breath alcohol concentration of a driver's exhaled breath, and more particularly to a method for quickly estimating a driver's breath alcohol level. [Background technology]

[0003] In an effort to prevent drunk driving, police regularly administer supervised breath tests. In addition to supervised breath tests, automatic ignition interlock devices (sometimes called "alcolocks") have been installed within vehicles themselves to perform unsupervised breath tests and prevent intoxicated drivers from operating the vehicle. The detection technologies used for such breath tests may be based on catalytic beads (or pellistors), semiconductors, fuel cells, or infrared spectroscopy. Fuel cells are the predominant detection element used for breath analyzers and alcolocks with mouthpieces. Evidential instruments typically use infrared spectroscopy. Catalytic beads, pellistors, and the like are typically used in low-cost devices for the consumer market and generally do not meet requirements for analytical specificity. Typical breath test devices provide a signal representing the breath alcohol concentration (BrAC) after the driver takes a deep breath and clears their airway into the mouthpiece, which is often separate and disposable for hygiene reasons. To ensure an accurate determination, the person being tested is required to deliver a forced expiration that approximates full vital capacity. This requires significant time and effort, especially for people with limited lung capacity. More specifically, it should be understood that there is a significant variation in what is considered "full vital capacity" when considering a typical population. Certain passive breath tests performed in accordance with the present invention are possible without obtaining a forced expiration that approximates full vital capacity for a given population (as would otherwise be required to perform an active breath test). It will be appreciated that the present invention generally utilizes a mid-expiration sample volume (e.g., 0.7-1.2 L) that is well below the full vital capacity of a typical adult.

[0004] Ease of use, convenience, and accuracy are important factors in increasing the acceptance and widespread use of self-contained ignition interlock devices in vehicles. Summary of the Invention [Problem to be solved by the invention]

[0005] Therefore, there is a need for a passive breath test that is flexible enough to avoid inconveniencing the driver while ensuring test accuracy under a wide range of environmental conditions and driver behavior. In a passive breath test, the driver is not required to provide air to be delivered to the sensor; the BrAC measurement is made from the air inside the vehicle, which is a mixture of both the driver's and any passenger's breath as well as ambient air, without any additional driver action. In contrast, an active breath test may require the driver to approach the sensor and deliver a forced, undiluted breath toward the sensor or through an air inlet (e.g., by blowing into a tube). While passive breath tests are preferred, some conditions may prevent accurate passive breath tests. These conditions may be environmental (e.g., very hot weather) or the result of the driver attempting to override the system (several examples are described below), but either way, the air inside the vehicle may not accurately reflect the driver's BrAC. When normal testing conditions for accurate passive BrAC testing are not met, active breath testing is required. [Means for solving the problem]

[0006] Various parameters indicative of both environmental conditions and driver behavior are measured to detect when normal test conditions are no longer met. These include, for example, detecting a peak in tracer gas concentration, indicating that the driver's breath has been detected. A timer can set a time limit between when the driver's presence is detected and when the peak in tracer gas concentration is detected. This time limit can prevent the driver from attempting to defeat the system by holding their breath or otherwise hiding their breath from the sensor. A pressure sensor can detect situations in which the driver is attempting to defeat the system by ventilating the vehicle, or situations in which wind blowing through the vehicle could interfere with an accurate passive breath test. Detecting the position of the driver's head relative to the sensor can ensure that the driver's breath is directed toward the sensor to prevent attempts to defeat the system by providing an alternative source of "breath" to be measured.

[0007] The methods and devices described herein enable passive detection of breath alcohol concentration and can be used to control vehicle ignition. Specifically, the methods and devices are designed to determine BrAC from a passive breath test without inconveniencing the driver while normal test conditions are met, and to detect when normal test conditions are no longer met and, under those circumstances, provide a BrAC measurement from an active breath test.

[0008] In one example of a method or device for passive breath alcohol detection for driving a vehicle, the device can include a sensor that measures the concentration of a tracer gas in a passively acquired first air sample, while the method can include activating a sensor system, passively acquiring the first air sample, and measuring the concentration of the tracer gas from the first air sample. The device can include a processor that uses the sensor system to determine a set of test conditions based in part on the first air sample, and the method can include the determining step. If the set of test conditions is within a normal range and a peak in the tracer gas concentration is detected, the method or processor measures the driver's BrAC from the first air sample. If the set of test conditions is outside the normal range or a peak in the tracer gas concentration is not detected, the method or processor requests an active second air sample from the driver and measures the driver's BrAC.

[0009] In some embodiments, the method includes measuring a time interval between activating the sensor system and detecting a peak in the tracer gas concentration. In some embodiments, the device includes a timer that measures the time interval. If the time interval exceeds a predetermined time limit, the method or processor determines that the set of test conditions is outside of a normal range. In some embodiments, the device includes a sensor that measures an environmental condition of the vehicle, and in some embodiments, the method includes measuring such an environmental condition. In some embodiments, such a sensor may be a temperature sensor, and in some embodiments, the method includes measuring the temperature inside the vehicle. If the temperature is outside of a normal temperature range, the method or processor determines that the set of test conditions is outside of a normal range. In some embodiments, the device includes a pressure sensor that measures the pressure inside the vehicle, and in some embodiments, the method includes measuring the pressure. If the pressure is outside of a normal pressure range, the method or processor determines that the set of test conditions is outside of a normal range. In some embodiments, the device includes a camera that measures the driver's head position relative to the BrAC sensor, and in some embodiments, the method includes measuring the driver's head position relative to the BrAC sensor. In some embodiments, activating the sensor system includes detecting the presence of a driver entering the vehicle. In some embodiments, measuring the driver's BrAC from an active breath test includes determining whether the BrAC measured from the passive breath test is at an intermediate level. If the measured BrAC is at an intermediate level, the method requests an active breath sample and measures the BrAC. In some embodiments, measuring the driver's BrAC from the active breath test includes requesting an undiluted breath sample via a human-machine interface (HMI) that is directed to a BrAC sensor.

[0010] In some embodiments, the method includes sending a sensor signal to a central processing unit (CPU) of the breath testing system in communication with the vehicle's CPU. In some embodiments, the processor receives the sensor signal from the sensor. In some embodiments, the method includes disabling operation of the vehicle if the driver's BrAC measurement results in a value above a set value, which may also be performed using the processor. In some embodiments, the method includes enabling operation of the vehicle if the driver's BrAC result is below a set value, which may also be performed using the processor. In some embodiments, the method includes requesting an active second air sample from the driver to measure the driver's BrAC if the passive breath test result of the first air sample is at an intermediate level. In some embodiments, the request may be performed by the processor. In some embodiments, the method includes continuously measuring air samples after activating the sensor system, and continuously measuring the concentration of the tracer gas after the first air sample is measured. In some embodiments, the sensor continuously measures air samples after activating the sensor system, and also continuously measures the concentration of the tracer gas after the first air sample is measured.

[0011] In some embodiments, the methods and devices are designed to accumulate sensor signal information over a series of exhalations until a desired level of confidence is achieved in the analyte measurement.

[0012] In a preferred form of the invention, there is provided a method for passive breath alcohol detection, the method comprising: A) passively obtaining a first air sample from air within an interior of a vehicle; B) determining the concentrations of (i) the tracer gas and (ii) the analyte present in the first air sample; C) passively obtaining a second air sample from the air within the interior of the vehicle; D) determining the concentrations of (i) the tracer gas and (ii) the analyte present in the second air sample; and E) continuing to passively obtain N air samples from the air within the interior of the vehicle, and for each air sample obtained, determining the concentrations of the tracer gas and the analyte present in the air sample; F) determining the number of tracer gas concentration peaks and the number of analyte concentration peaks present in each of the air samples; G) determining a confidence interval based on the number of tracer gas concentration peaks and the number of analyte concentration peaks; H) controlling operation of the vehicle based on the confidence interval and a function of the concentration of the analyte present in the air sample.

[0013] Further features of the present subject matter, its nature and various advantages will become apparent upon consideration of the following detailed description taken in conjunction with the accompanying drawings, in which like reference characters refer to like parts throughout. [Brief explanation of the drawings]

[0014] [Figure 1] 1 is a flowchart of a process for determining a BrAC measurement from a passive or active breath sample, according to an exemplary implementation. [Figure 2] 1 is a flowchart of a process for determining the results of a BrAC measurement from a passive breath sample, according to an exemplary implementation. [Figure 3] 1 is a flowchart of a process for determining the results of a BrAC measurement from an active breath sample, according to an exemplary implementation. [Figure 4] FIG. 1 illustrates a sensor for detecting exhaled breath and BrAC concentration from both active and passive breath samples, according to an exemplary implementation. [Figure 5A] 1 illustrates an overhead view of a driver's head position relative to sensors according to an exemplary implementation. [Figure 5B] 1 illustrates an overhead view of a driver's head position relative to sensors according to an exemplary implementation. [Figure 6] 10 is a graph illustrating an example of signals detected by a wake-up sensor and a BrAC sensor according to an exemplary implementation. [Figure 7] 10 is a flow chart illustrating one method for accumulating sensor signal information over a series of exhalations until a desired confidence level is achieved in an analyte measurement. [Figure 8] 10 is a flow chart illustrating another method for accumulating sensor signal information over a series of exhalations until a desired confidence level is achieved in an analyte measurement. [Figure 9] 10 is a graph illustrating an example of signals detected by a wake-up sensor and a BrAC sensor according to another exemplary implementation. DETAILED DESCRIPTION OF THE INVENTION

[0015] background Breath testing of drivers is an effective screening method to reduce drunk driving and drunk driving-related deaths. In a breath test, a subject exhales air into a sensor or measuring device for a sufficient time and volume to achieve expiratory flow through the alveoli of the lungs, where substances such as ethyl alcohol (EtOH) in the blood are exchanged with air. The sensor or measuring device then measures the alcohol content (BrAC) of the air, which is related to blood alcohol by a conversion algorithm.

[0016] Existing breath-based alcohol testing technologies require drivers to forcefully exhale nearly to their lung capacity, which typically requires significant time and effort, especially for people with limited lung capacity. For hygiene reasons, the mouthpieces used in existing breath testing devices also need to be cleaned and replaced after multiple uses. Additionally, environmental conditions such as wind, temperature, and the presence of other people can significantly affect the accuracy of BrAC measurements. To promote widespread adoption and social acceptance of vehicle ignition interlock devices, a breath testing system is needed that does not inconvenience drivers and is robust under a wide range of conditions found in vehicles.

[0017] Therefore, there is a need for a passive breath test that is flexible enough to avoid inconvenience to the driver while ensuring test accuracy under a wide range of environmental conditions and driver behavior. In a passive breath test, the driver does not need to provide air to be delivered to the sensor; BrAC measurements are made from the air inside the vehicle, which is a mixture of both the driver's breath and any passengers' breath as well as ambient air, without any additional driver action. In a passive breath test, the air inside the vehicle is drawn into the sensor using a fan. BrAC measurements are made by first measuring the concentration of a tracer gas, such as carbon dioxide, which indicates the presence of the driver's breath in the air inside the vehicle. The measured EtOH concentration can then be combined with a dilution factor for this breath to determine BrAC. In this way, BrAC measurements are made without inconvenience to the driver by simply sampling the air inside the vehicle.

[0018] In contrast, in an active breath test, the driver may be required to approach a sensor and forcefully expel undiluted breath toward the sensor or through an air inlet (e.g., by blowing through a tube). In an active breath test, BrAC is thus measured directly from the driver's breath rather than through the air inside the vehicle. Active breath tests require action by the driver in addition to the normal action required to start the vehicle and may therefore be considered less convenient than passive breath tests.

[0019] Although passive breath testing is preferred, some conditions may prevent accurate passive breath testing. For example, the driver may attempt to disable the system, resulting in the air inside the vehicle not accurately reflecting the driver's BrAC. Similarly, environmental conditions inside the vehicle (such as strong drafts through an open window or high temperatures after the vehicle has been closed in hot weather) may prevent accurate BrAC measurements. When the normal test conditions for accurate passive BrAC testing are not met, the driver will be required to take an active breath test.

[0020] The present invention provides various detection checks to allow passive detection and estimation of a driver's BrAC under normal conditions, and switches to BrAC measurement from an active breath test when normal conditions are no longer met. This reduces inconvenience to the driver by defaulting to passive estimation of BrAC, while simultaneously providing an alternative logic path when the accuracy of the BrAC estimate is not reliable or when driver behavior or test conditions deviate from norms.

[0021] BrAC measurement from passive or active breath samples FIG. 1 shows a flowchart of a process for determining a BrAC measurement from a passive or active breath sample according to one exemplary implementation. Process 100 begins at 102. Start 102 may be initiated by a wireless door key unlocking the vehicle door, the opening of the driver's seat door, or any other indicator that the driver has entered the vehicle. When the driver or test subject takes their first step toward entering the driver's seat of the vehicle, process 100 will activate sensors during the test at 104 to monitor test conditions within the vehicle, detect the driver's breath, and start a timer to verify whether the driver's breath is detected within a time limit. If the driver's breath is detected (as determined by logic gate 106), the test conditions within the vehicle are normal (as determined by logic gate 108), and the time limit has not been exceeded (as determined by logic gate 110), process 100 will proceed to 112, where a BrAC measurement will be made from the passive breath sample. However, if either logic gate 106 or logic gate 108 returns a negative determination or the time limit has been exceeded at 110, process 100 will proceed to request an active breath sample from the driver at 114.

[0022] Testing at 104 includes self-testing of all functional blocks and sensors used in process 100. Testing 104 establishes stable operating temperatures for temperature sensitive elements of any sensors used in process 100.

[0023] This may include, for example, heating a mirror in the tracer gas detection sensor to above 40°C. The mirror and tracer gas detection sensor are described in further detail with reference to FIG. 4. The self-test procedure performed at 104 may last between 5 and 8 seconds at test conditions at or above room temperature. At lower temperatures, the self-test procedure may last longer than 8 seconds. The test at 104 may also involve measuring the initial conditions of the vehicle before the driver enters, such as the CO2 level, EtOH concentration, air temperature, and air pressure of the vehicle's air. These initial conditions can be used to determine whether the driver's breath is being detected at logic gate 106 and whether the test conditions are within normal ranges at logic gate 108.

[0024] The tracer gas can be any gas used to detect the driver's exhaled breath. The tracer gas can be carbon dioxide (CO2) or any other gas capable of indicating exhaled breath. The sensitivity of the tracer gas detection sensor allows for detection of highly diluted exhaled breath, where the dilution factor (i.e., the ratio of ambient air to undiluted exhaled breath) can be 50 or greater. Following initiation of process 100, air is continuously drawn from the air inside the vehicle through the tracer gas detection sensor. The tracer gas detection sensor can be positioned closer to the driver's head position than any passenger position, for example, on the steering column or on the side door closest to the driver's side of the vehicle. The exhaled breath is recognized by the tracer gas detection sensor as a peak output signal. If the tracer gas is CO2, the reference concentration of CO2 corresponding to the reference signal is expected to be between 400 and 600 ppm (0.04% to 0.06% by volume). The tracer gas signal and activation signal at 102 are described in more detail below with reference to Figure 6. The tracer gas detection sensor determines whether breath is detected at logic gate 106, which is described in more detail below with reference to Figure 4. The breath sample in which the tracer gas detection peak is observed may be the same breath sample used to measure BrAC at 112. Thus, detection of the driver's breath during test 104 and logic gate 106 can occur substantially simultaneously with the BrAC measurement from the passive breath sample made at 112.

[0025] Logic gate 108 can process the vehicle's test conditions and determine whether the test conditions fall within the normal conditions that allow accurate BrAC measurements to be generated from passive breath samples.

[0026] Environmental conditions may involve both driver behavior and the state of the vehicle itself. These conditions can be detected by various sensors, including tracer gas detection sensors and auxiliary sensors located throughout the vehicle. Sensors may include temperature sensors that determine the temperature inside the vehicle and pressure sensors that determine the air pressure inside the vehicle and the wind or air flowing through the vehicle. Normal temperatures inside a vehicle may range from -40°C to 85°C. This may be the temperature range within which a BrAC sensor can perform an accurate passive breath test. Normal air pressure may range from 80 to 105 kPa. This may be the pressure range within which the mixture of the driver's breath and ambient air can provide an accurate passive breath test. The temperature and pressure sensors may be any standard sensor elements and may be integrated into the vehicle body.

[0027] A camera sensor monitoring the driver's behavior may also be positioned near the driver, such as near the steering wheel column. As described in more detail below with respect to FIGS. 5A and 5B, this camera sensor can detect the position of the driver's head relative to the tracer gas detection sensor. The relationship between the driver's head and the tracer gas detection sensor can determine whether the driver is breathing in the direction of the tracer gas detection sensor. The camera can detect scenarios in which the driver is trying to avoid detection of their BrAC by turning their face away from the sensor. The camera sensor can also detect the presence of unfamiliar objects near the driver's face, such as a mask, filter, spray bottle, or other object that attempts to interfere with the tracer gas detection sensor or provide an alternative "breath" source, preventing an accurate passive breath test. The camera sensor can also detect the proximity posture of the occupant, which may make it difficult to distinguish between the occupant's BrAC level and the driver's BrAC level, or in which the driver is trying to have the system measure the occupant's BrAC level instead of their own.

[0028] Logic gate 108 can also determine the status of the vehicle's heating, ventilation, and air conditioning (HVAC) system, such as whether it is on or off. Preferably, the HVAC system is turned off or at normal operating conditions during process 100. Using the HVAC system during a passive breath test could overly dilute the driver's EtOH level, divert the driver's breath from the sensor, or otherwise interfere with an accurate passive breath test. Logic gate 108 can also detect the presence of windshield fluid. Windshield fluid typically contains ethyl alcohol, which can affect the detection of EtOH in the vehicle. Windshield fluid is off under normal test conditions. Logic gate 108 can determine the status of both the vehicle's HVAC system and windshield fluid by communicating with the vehicle via the vehicle's CPU or a controller area network (CAN) bus, for example.

[0029] A logic gate at 110 determines whether the time limit for process 100 to detect the driver's breath during test 104 has been exceeded. This may be a predetermined time limit, such as 10-30 seconds. If two or more tracer peaks are detected within the time limit, the average and difference between each BrAC measurement can be used to add confidence to the classification into classes "high," "medium," and "low," as described further below. The accumulated tracer concentration is a key factor in adding confidence by increasing the accumulated signal-to-noise ratio. More specifically, if two or more peaks of tracer gas concentration are detected within the time limit, the average and difference between each BrAC measurement (i.e., the change in the area under each peak for successive BrAC concentration signals) can be used to add confidence to the classification of the accuracy of the BrAC measurement. By way of example and not limitation, as discussed in more detail below, a set of cumulative BrAC concentration measurements taken over a predetermined time period can be used to calculate a driver's BrAC, with the reliability of the calculated BrAC being classified into classes "high reliability," "medium reliability," and "low reliability." The cumulative tracer gas concentration (i.e., determined from multiple tracer gas concentration measurements) is a key factor in adding reliability by increasing the cumulative signal-to-noise ratio. If this time limit is determined to have been exceeded at 110, process 100 proceeds to measure BrAC from an active breath sample at 114. The time limit can prevent situations where the driver is avoiding breathing toward the sensor, holding their breath, wearing a mask on their head, or otherwise attempting to operate the vehicle without providing a breath sample. In this case, the logic gate at 110 will recognize that no breath has been detected after the predetermined time limit and will request an active breath sample from the driver at 114.

[0030] If exhaled driver breath is detected at logic gate 106, logic gate 108 determines that the test conditions are normal, and logic gate 110 determines that the time limit has not been exceeded, process 100 will proceed to measuring BrAC from a passive breath sample at 112. BrAC measurement 112 may be described in further detail with reference to FIG. 2. However, if no breath is detected at 106, process 100 will continue to test for driver breath until either the time limit is exceeded at 110 or the normal test conditions are no longer met at 108. In this case, process 100 will request an active breath sample at 114 in which BrAC is measured. A determination at 108 that the normal test conditions are not met is sufficient to request an active breath sample at 114. Similarly, if the time limit is exceeded as determined at 110, process 100 will proceed to an active breath sample at 114. The active breath sample at 114 may be described in further detail with reference to FIG. 3.

[0031] The results of the BrAC measurements 112 and 114 may differ in accuracy. BrAC measurement from passive breath samples FIG. 2 shows a flowchart of a process for determining a BrAC measurement from a passive breath sample according to one exemplary implementation. Process 112 determines the driver's BrAC at 202. Process 112 may be substantially simultaneous with the detection of tracer gas at 104 and 106 shown in FIG. 1. Thus, process 112 may be performed using the same breath sample collected at 104 to detect a tracer gas peak indicative of the driver's breath. In other words, if initiation of process 100 causes breath to be detected at logic gate 106 within the time limit established at 104, process 112 may proceed to determine the driver's BrAC using the first breath sample recognized at logic gate 106. The measurement of BrAC is based on the dilution of the detected tracer gas, which may be CO2. A detected dilution factor, or DF, of the tracer gas in the vehicle's ambient air is used to determine an estimate of the driver's BrAC. The BrAC level may be determined according to the following equation: BrAC=EtOH*DF (Formula 1) DF is the dilution factor of the tracer gas in the air; i.e., DF is the ratio of the end expository (undiluted) tracer gas concentration measured by the sensor to the tracer gas concentration. Additional algorithms incorporating information from auxiliary sensors (not shown) may be used. The algorithm used to measure BrAC from passive exhaled breath samples at 112 is essentially the same as the algorithm used to measure BrAC from active exhaled breath samples at 114. It will be appreciated that the DF is much greater when measuring BrAC from passive exhaled breath samples at 112 than when measuring BrAC from active exhaled breath samples at 114. As seen in FIG. 2, if the estimated BrAC is below a predetermined set value (labeled "low" or "L"), process 112 outputs a signal at 206 that the driver's BrAC is "fair." The predetermined set value may range from 0.1 to 0.4 mg / L (50 to 200 ppm). The predetermined set point may be a function of the driver's age. The predetermined set point may also be a function of the legal limit imposed on blood alcohol concentration for driving under the influence (DUI) or driving while impaired (DWI). Signal 206 may be used to enable operation of the vehicle. If the estimated BrAC value is significantly above the predetermined set point (denoted "high" or "H"), process 112 will output a signal at 204 indicating that the driver's BrAC is "unacceptable." By way of example and not limitation, a BrAC level 0.1-0.2 mg / L (50-100 ppm) above the set point may cause the output of a signal at 204 indicating that the driver's BrAC is "unacceptable." Signal 204 may be used to disable operation of the vehicle. If the driver's BrAC is determined to be in the intermediate range (labeled "Intermediate" or "I"), slightly above or below a predetermined set point, at 202, further analysis of the driver's breath is required to make a final determination. The driver is then asked to take an active breath test at 208.Active breath testing is described in further detail with reference to Figure 3 and is process 114 shown in Figures 1 and 3. A sensor such as that described in Figure 4 can take the measurement at 202 and external processing as described in more detail in Figure 4 can convey signals 204 and 206 to a central processing unit or CPU 415.

[0032] BrAC measurement from active breath samples 3 shows a flowchart of a process for determining a BrAC measurement from an active breath sample according to one exemplary implementation. Process 114 begins at 302, where the driver's BrAC is measured. To measure BrAC from an active breath sample at 302, the driver is asked to provide an active breath toward a sensor (not shown) at a distance of 15-30 cm from the sensor. The distance may be adjusted depending on the location of the sensor within the vehicle. If the BrAC is below a predetermined set value (denoted "low" or "L"), process 114 outputs a signal at 306 that the driver's BrAC is "fair."

[0033] Signal 306 may be used to enable vehicle operation. If the estimated BrAC value is significantly above a predetermined set value (denoted "high" or "H"), process 114 will output a signal at 308 indicating the driver's BrAC is "unavailable." Signal 308 may be used to disable vehicle operation. If the driver's BrAC is determined at 302 to be in the intermediate range (denoted "intermediate" or "I"), slightly above or below the predetermined set value, the driver will be requested to provide a breath sample with evidential accuracy at 304. The breath test performed at 304 will require an undiluted breath sample. There will be no intermediate response from test 304. The breath test at 304 will require the driver to make an active exhalation toward a sensor (not shown) at a distance of 15-30 cm from the sensor. The distance may be adjusted depending on the location of the sensor within the vehicle. If the measured BrAC value is below the set value (L), the process 114 will generate an output signal indicating that the driver's BrAC is "passable" at 306. If the measured BrAC value is above the set value (H), the process 114 will generate an output signal indicating that the driver's BrAC is "fail" at 308.

[0034] Although logic gates 106, 108, 110 of Figure 1, logic gate 202 of Figure 2, and logic gates 302 and 304 of Figure 3 are shown as representing "IF...THEN" logic statements, these logic gates are not limited to elements of Boolean logic. Logic gates 106, 108, 110, 202, 302, and 304 may also represent fuzzy logic or be controlled by an artificial neural network. As will be described in more detail with reference to Figure 4, the desired logic system may be programmed into CPU 415.

[0035] Sensor for detecting breath and BrAC concentrations from both active and passive breath samples FIG. 4 illustrates a sensor for detecting breath and BrAC concentrations from both active and passive breath samples according to one exemplary implementation. Sensor 400 detects breath during test 104 shown in FIG. 1 , measures the vehicle's initial conditions at 104, and is capable of making BrAC measurements from both passive and active breath samples 112 and 114, as shown in FIG. 1 and described in further detail with reference to FIGS. 2 and 3 . Sensor 400 continuously draws air through inlet 402, passes the air to outlet 403, and measures both the presence of a tracer gas, such as CO, and the presence of EtOH in the airstream. As described with reference to FIGS. 2 and 3 , sensor 400 determines whether the driver's BrAC is in the low, high, or intermediate range.

[0036] The sensor 400 is housed in a housing 401 and may be a stand-alone sensor or may be designed to be integrated into the interior of the vehicle, such as on the steering wheel column, a side door, a vertically supported A-pillar or a vertically supported B-pillar, a sun visor, the dashboard, or another convenient location significantly closer to the driver's head than the area designated for vehicle occupants. The housing 401 may be airtight except for the sensor inlet 402 and outlet 403 openings. The approximate dimensions of the housing 401 may be 25 x 40 x 120 mm. Air drawn into the housing 401 through the inlet 402 is heated above body temperature by an inlet heater 404, which may prevent condensation at low ambient temperatures. The inlet heater 404 may have a large contact surface area with the intake air to improve heat transfer from the heater to the incoming air. The heater 404 may be a resistive heater. The airflow from the inlet 402 to the outlet 403 is driven by a fan 411 positioned near the outlet 403.

[0037] Sensor 400 measures the presence of both CO2 and EtOH by infrared (IR) spectroscopy. IR spectroscopy determines the alcohol concentration in the air stream of sensor 400 using the unique "fingerprint" that vapor-phase alcohol produces when illuminated by infrared light. The detected absorption spectrum of any substance is the product of resonant molecular vibrations, which are unique to a molecule or atomic bond within a compound in the breath sample. From the absorption spectrum, specific substances and their absolute or relative concentrations in the breath sample can be determined.

[0038] To perform IR spectroscopy and detect both the presence of the tracer gas and the presence of EtOH, the sensor air chamber tube 410 includes two separate optical paths, one for detecting the tracer gas and one for detecting EtOH. The signals generated by these two optical paths are used to determine the value of the dilution factor of the driver's breath in the ambient air (or DF shown in Equation 1) and the value of the EtOH concentration in the input air.

[0039] 4, a first optical path consisting of an EtOH IR emitter 406 and an EtOH IR detector 407 will determine the EtOH concentration. The EtOH IR emitter 406 outputs IR radiation into the interior of a sensor air chamber tube 410. A spherical mirror assembly consisting of a first mirror 405 located at one end of the air chamber tube 410 and a second mirror 412 located at the opposite end of the air chamber tube 410 reflects the IR radiation emitted from the EtOH IR emitter 406. Because the mirror assembly will reflect the emitted light multiple times before it hits the EtOH IR detector 407, the optical path length of the emitted IR radiation can be several times the distance between the first mirror 405 and the second mirror 412. Mirrors 405, 412 preferably include a layer of highly reflective material, such as gold or aluminum, and a very thin protective surface layer on their reflective surfaces to maintain high reflectivity even when air chamber tube 410 is subjected to corrosive gases during its operational life. The optical path is shown by a dashed line in Figure 4, but this is shown by way of example and the actual optical path between EtOH IR emitter 406 and EtOH IR detector 407 may be much longer. Typically, the beam of infrared radiation may traverse the optical cell from emitter 406 multiple times, for example, 16 times (or a multiple of 4) before the beam hits detector 407.

[0040] The EtOH IR emitter 406 may be a black-body radiator, an IR laser diode, or any other light source capable of generating IR light and preferably small enough to fit within the air chamber tube 410. The EtOH IR emitter 406 may be modulated at a frequency between 5 and 10 Hz to suppress low-frequency noise and disturbances in the signal from the EtOH IR detector 407. The EtOH IR detector 407 includes a bandpass filter tuned to the IR absorption peak of EtOH, which is approximately 9.5 μm. The EtOH IR detector 407 may be a pyroelectric or photon detector capable of generating a high-resolution signal and may also include a Peltier element for local cooling to suppress thermal noise in the detected signal. The detected signal generated by the EtOH IR detector 407 is described in further detail with reference to FIG. 6. Sensor 400 is specifically designed for high-resolution IR spectroscopy measurements and can have a resolution better than 0.5 μg / L (or 1.2 ppm) of EtOH, allowing estimation of alcohol concentrations in highly diluted breath.

[0041] The second optical path is intended to detect the presence of a tracer gas, such as CO2, which would indicate the driver's exhaled air being mixed into the air input through sensor 400. Tracer gas IR emitter 408 is positioned opposite tracer gas IR detector 409, such that the optical path from tracer gas IR emitter 408 to tracer gas IR detector 409 spans the shorter dimension of air chamber tube 410. Tracer gas IR detector 409 may be tuned to a wavelength band specific to the IR absorption frequency of the tracer gas being detected. In one example where the tracer gas is CO2, the absorption peak may be at 4.26 μm. Because the end-tidal concentration of CO2 in exhaled air is high (typically 4.2% by volume), a short optical path across air chamber tube 410 may be used. This path is shown in FIG. 4 as the dashed line between tracer gas IR emitter 408 and tracer gas IR detector 409. The signal produced by the IR detector for the tracer gas is described in further detail with reference to FIG.

[0042] The signals from both the EtOH IR detector 407 and the tracer gas IR detector 409 can be used to determine whether normal conditions for passive measurement of BrAC are met, such as during test 104 and logic gate 108 of process 100 shown in Figure 1. This process may involve verifying that a baseline presence of the tracer gas, such as a baseline concentration of CO2 in ambient air, is within a standard level.

[0043] Both the first mirror 405 and the second mirror 412 are in communication with a central processing unit or CPU 415 shown in Figure 4. The EtOH IR emitter 406, the EtOH IR detector 407, the tracer gas IR emitter 408, and the tracer gas IR detector 409 may also be in signal communication with the CPU 415. The CPU 415 processes signals generated by the EtOH IR detector 407 and the tracer gas IR detector 409 to determine a BrAC measurement.

[0044] The human-machine interface (HMI) 413 is in communication with the CPU 415 and may be used to communicate with the driver to request a BrAC measurement from an active breath sample. The HMI 413 may include audiovisual means for driver interaction, such as a screen and speaker, to convey messages, requests for an active breath test, and other instructions to the driver. The HMI 413 may display the results of the BrAC measurement to the driver. The HMI 413 may be a multi-purpose interface, such that requesting and displaying information about the breath test is only one of many functions. Other functions may be navigation, interaction with the HVAC, interaction with the stereo system, or interaction with other systems typical of a vehicle. The HMI 413 may be integrated into the vehicle within the driver's field of view.

[0045] The CPU 415 is also in communication with an auxiliary sensor 414, which may be, for example, a temperature, barometric, or optical sensor, or a camera, that determines test conditions within the vehicle. The auxiliary sensor 414 is used during test 104 of process 100. A data communication unit 416 can store BrAC measurements and parameter values ​​used by the CPU to determine normal test conditions of the vehicle. The data communication unit 416 also transfers data between the sensor system 400 and other units (not shown) external to the sensor 400. The sensor 400 also includes a power unit 117 for managing and providing power.

[0046] Measuring head position FIG. 5A illustrates an overhead view of the driver's head position relative to the sensor, according to an exemplary implementation. The auxiliary sensor measures a lateral distance 510 between the driver 502 and the sensor 506. The lateral distance 510 may be measured between the midline 508 of the sensor 506 and the midline 504 of the driver's head 502. Under normal test conditions, the lateral distance 510 is 20 cm or less. The measurement of the lateral distance 510 may be performed in the test 104 of the process 100 shown in FIG. 1. The determination in the logic gate 108 of whether the vehicle's test conditions are normal takes into account the lateral distance 510 shown in FIG. 5A. The measurement of the lateral distance 510 can ensure that the driver is breathing directly toward the sensor 506 during an active breath test. The measurement of the lateral distance 510 can also ensure that the driver is not trying to avoid the sensor's breath detection by moving their head away from the sensor during a passive breath test.

[0047] 5B, the auxiliary sensor also measures the rotation of the driver's head relative to the midline 518 of the sensor 516. A rotation angle 520 may be measured between the midline 518 of the sensor 516 and the midline 514 of the driver's head 512. Under normal test conditions, the rotation angle 520 may be ±5°. The measurement of the rotation angle 520 may be performed in test 104 of the process 100 shown in FIG.

[0048] The determination at logic gate 108 of whether the vehicle's test conditions are normal or not typically takes into account rotation angle 520, shown in Figure 5B. Measuring rotation angle 520 can ensure that during an active breath test, the driver is breathing directly towards sensor 516. Measuring rotation angle 520 can also ensure that during a passive breath test, the driver is not trying to avoid the sensor's breath detection by turning their head away from the sensor.

[0049] The lateral distance 510 and rotation angle 520 may be measured by an integrated camera sensor (not shown), which may be integrated into the vehicle and positioned near the driver's head. The integrated camera may also determine whether an unfamiliar object is within the camera's field of view and / or whether an occupant is within the driver's head field of view.

[0050] Signals detected by the activation sensor and BrAC sensor FIG. 6 is a graph illustrating an example of a signal detected by sensor 400 using one exemplary implementation of the present invention shown in FIG. 4. The graphs shown at 600 occur along the same time scale. Graph 602 (labeled "P") shows signal 604 representing the detection of a driver's presence (e.g., detected by auxiliary sensor 414) when the driver enters or prepares to enter the vehicle. Signal 604 initiates start 102 shown in process 100 in FIG. 1. Signal 604 starts a timer that determines the time interval between the activation signal at 604 and the detection of exhalation shown in graph 606 (labeled "T"). Exhalation is detected from a peak 608 in the tracer gas concentration signal. When the tracer gas is CO2, the peak exhalation detection is characterized by a CO2 detection magnitude of 525 ppm or greater, assuming a dilution factor of 80 and an end-tidal CO2 concentration of 4.2% by volume. The duration 614 of the peak 608 is expected to be 1-3 seconds, excluding the sensor response time. Additionally, the rise time 612 and fall time 616 of the peak can be used to characterize the tracer gas signal resulting from the driver's exhaled breath. A typical rise time can be in the range of 0.5-1.0 seconds, and a typical fall time can be in the range of 3-5 seconds. The peak of the tracer gas signal at 610 can terminate a timer. A time interval 622 can be measured and used to determine whether an acceptable time limit has been exceeded between the activation signal 604 and point 610 of the exhaled breath signal 608, such as in logic gate 110 shown in FIG. 1 .

[0051] Graph 618 (labeled "A") shows the detected signal for EtOH. Depending on the concentration of EtOH in the driver's breath, graph 618 may or may not show peak 620 corresponding to tracer gas peak 608. However, if EtOH is present in the driver's breath, EtOH signal 620 will be approximately simultaneous with tracer gas signal 608, as shown in FIG. 6. The magnitude of the peak in EtOH signal 620 will indicate the concentration of BrAC in the measured breath sample. Measurement of BrAC from peak 620 can occur approximately simultaneously with detection of peak 608 in the trace gas signal.

[0052] Passive BrAC Estimation Using Cumulative Sensor Signal Acquisition In the above sections, estimation of BrAC is performed through the combined use of sensor signals indicative of tracer gases, e.g., carbon dioxide (CO2) and ethyl alcohol vapor (EtOH). However, the present invention is not limited to these substances or to the specific location of the test subject (e.g., the driver in the driver's seat). The present invention may be used in any situation where it is critical to accurately estimate the breath concentration of any particular substance without interfering with the test subject.

[0053] As mentioned above, a passive estimation of breath alcohol concentration (BrAC) can be performed by measuring the tracer gas concentration at the same location and performing the following calculation: BrAC=EtOH * DF (Equation 1) Here, DF denotes the dilution factor, which is determined by dividing the end-tidal concentration of the tracer gas by the measurement at the sensor location. For CO2 as the tracer gas, the end-tidal concentration is 4.2% by volume, and the corresponding value for water vapor is 5.5% by volume. In passive in-vehicle applications, DF can vary considerably.

[0054] The above sections disclose methods and devices for passive detection of analytes (e.g., EtOH) by controlling conditions related to signal stability, environmental influences, and subject behavior, all of which are necessary for accurate analyte estimation. When these conditions deviate from normal, the accuracy of the passive detection process may decrease, and the subject may be required to provide an active breath sample.

[0055] In the following sections, improved methods and devices for passive detection of analytes (e.g., EtOH) are disclosed that can provide more accurate estimations of analytes (e.g., EtOH). The improved methods and devices are essentially based on the premise that increasing the amount of gas analyzed increases the confidence level of the analyte assessment, i.e., accumulating sensor signal information over a series of exhalations can increase the confidence level of the analyte measurement.

[0056] Some of the key features of this improved method and apparatus are: - a means for determining and repeatedly recording instantaneous tracer gas concentrations at a location near the driver's seat of the vehicle; - means for determining and repeatedly recording instantaneous alcohol vapor concentrations synchronously with and at the same location as the recording of tracer gas concentrations; - a means of quantifying the magnitude and timing of the tracer gas concentration peak; - a means for calculating cumulative breath alcohol concentration based on the magnitude and timing of the tracer gas peaks (i.e., by using instantaneous tracer gas and alcohol vapor concentrations); - a means for calculating a cumulative confidence level of breath alcohol concentration over time; and - means for activating an alarm and / or enabling and / or disabling the operation of the vehicle according to the combined result of the average breath alcohol determination and the calculation of the confidence level in relation to a preset limit concentration; Includes.

[0057] Essentially, this aspect of the invention recognizes that the confidence level of an analyte determination increases as the amount of gas analyzed increases, i.e., the confidence level of an analyte determination increases as the number of breaths analyzed increases.

[0058] Referring now to FIG. 7, a signal flow graph corresponding to one embodiment of an improved method and apparatus for estimating BrAC is shown. The system begins at 701 by unlocking the vehicle cabin door, and a startup phase 702 includes a self-test procedure that checks the operation of all system components. These procedures are typically quick, allowing the system to quickly continuously record sensor signals corresponding to (i) local tracer gas concentration 703 and (ii) EtOH concentration 704. When the driver enters the vehicle cabin and expels exhaled breath, a tracer gas peak will appear at the sensor location, and the magnitude and timing of this peak will be determined by delta detector 705. The timing is used to trigger the determination of the magnitude of the corresponding EtOH peak by delta detector 706. A more detailed description of the function of delta detectors 705, 706 is provided below in connection with FIG. 9.

[0059] The tracer gas and EtOH signals 703, 704 are continuously recorded. As time progresses, more signal peaks corresponding to exhaled breath from the driver will be recorded and accumulated by adding the contribution from each detected peak by summer block 707 for the tracer gas signal and summer block 708 for the EtOH signal.

[0060] The adder block signals 707, 708 are combined to allow for cumulative BrAC calculation 709 using the formula:

[0061]

number

[0062] The formula is:

[0063]

number

[0064] By both Equation 2 and Equation 3, the CO2-dependent dilution factor (DF) is a quality indicator. Thus, by both Equation 2 and Equation 3, it will be understood that the smaller the value of the dilution factor (DF) that is CO2-dependent, i.e., corresponds to a higher measured CO2 value, the more heavily the corresponding peak is weighted in Equations 2 and 3.

[0065] The confidence level of the BrAC value is then checked in block 710 with respect to (i) a criterion based on a legal limit value or any other preset limit value (e.g., a BrAC value of 0.08% or less may be required to operate a vehicle in the United States, a different BrAC value may be required to operate a vehicle outside the United States, etc.), and (ii) a required confidence level of the calculated cumulative BrAC, which may vary from one application to another (e.g., a “high confidence level” may be required when the BrAC value is very close to the legal limit value for a given jurisdiction (e.g., slightly above or slightly below the legal limit value), a “medium confidence level” may be required when the BrAC value is greater than zero but still significantly below the legal limit value for a given jurisdiction, a “low confidence level” may be required when the BrAC value is significantly below (significantly above) the legal limit value for a given jurisdiction, etc.). If the desired confidence level for a particular application (e.g., "high confidence level," "medium confidence level," or "low confidence level") is achieved, a final lock / unlock function 711 is reached that determines the vehicle's drivability. If the desired confidence level is not achieved, additional accumulations of BrAC determinations 709 will be required (i.e., the system will continue to acquire BrAC information from additional driver exhalations until the desired confidence level is achieved). It should be noted that various factors may affect the confidence level associated with the determination of analyte concentration, including, but not limited to, sensor sensitivity, number of exhalations sampled, signal stability, environmental effects, and subject behavior. It should also be noted that the confidence level ascribed to the accumulated BrAC value is a function of the number of exhalations analyzed to arrive at the accumulated BrAC value, which in turn is indicated by the number of tracer gas concentration peaks measured.

[0066] In another aspect of the invention, and referring now to Figure 8, the implementation of confidence block 809 is somewhat different and is based solely on the tracer gas signal. In this implementation of the invention, and as discussed in more detail below, BrAC determination 810 is only performed when the accumulation of tracer peaks reaches a certain preset limit based on the results of tests performed on the system. Lock / unlock function 811 is identical to lock / unlock function 711.

[0067] Thus, in this form of the invention, and again with reference to FIG. 8, the system begins at 801 by unlocking the vehicle cabin door, and a start-up phase 802 includes a self-test procedure that checks the operation of all system components. These procedures are typically quick, allowing the system to quickly continuously record sensor signals corresponding to local tracer gas concentration 803 and EtOH concentration 804. When the driver enters the vehicle cabin and expels exhaled breath, a tracer gas peak will appear at the sensor location, and the magnitude and timing of this peak will be determined by delta detector 805. The timing is used to trigger a determination of the magnitude of the corresponding EtOH peak by delta detector 806. Again, a more detailed description of the function of delta detectors 805, 806 is provided below in connection with FIG. 9.

[0068] The tracer gas and EtOH signals 803, 804 are continuously recorded. As time progresses, more signal peaks corresponding to exhaled breath from the driver will be recorded and accumulated by adding the contribution from each detected peak by summer block 807 for the tracer gas signal and summer block 808 for the EtOH signal.

[0069] The confidence of the tracer gas value is then checked in block 809 against (i) a standard based on legal limits or any other pre-set limits (e.g., a BrAC value of 0.08% or less may be required to operate a vehicle in the United States, or different BrAC values ​​may be established by law for different jurisdictions, or different BrAC values ​​may be set by the vehicle manufacturer, etc.), and (ii) a desired confidence level (e.g., a "high confidence level," a "medium confidence level," or a "low confidence level"). If the desired confidence level is achieved, adder blocks 807, 808 output a signal based on the formula: BrAC=EtOH * The DFs are combined to enable a cumulative BrAC calculation 810. Note that Equations 2 and 3 may be used when it is desired to attribute a higher statistical weight to the cumulative BrAC for measurements where higher peaks of tracer gas (CO2) concentration are present compared to measurements where lower peaks of tracer gas (CO2) concentration are present. It will be appreciated that when calculating the cumulative BrAC value, a higher statistical weight should be attributed to measurements with higher tracer gas concentrations compared to measurements with lower tracer gas concentrations (i.e., because an EtOH concentration signal corresponding to a higher tracer gas concentration signal is more likely to accurately indicate breath in the room compared to an EtOH concentration signal corresponding to a lower tracer gas concentration signal (which may indicate partial / over-diluted breath in the room)). If the required confidence level is achieved, a final lock / unlock function 811 is reached, which determines vehicle drivability. If the required confidence level is not achieved, additional accumulations of the BrAC determination 810 will be required (i.e., the system continues to acquire tracer gas information from additional driver exhalations until the desired confidence level is achieved). It should be noted that various factors may affect the confidence level associated with the determination of analyte concentration, including, but not limited to, sensor sensitivity, number of breaths sampled, signal stability, environmental influences, and subject behavior.

[0070] FIG. 9 shows a schematic representation of typical tracer gas and EtOH concentration signals as a function of time. These tracer gas and EtOH concentration signals correspond to the outputs from blocks 703, 803, and 704, 804, respectively, in FIGS. 7 and 8. As can be seen in FIG. 9, two peaks are observed in both the tracer gas and EtOH signals approximately 3 and 8 seconds after the starting point. The peaks correspond to exhaled breath from the subject. Note that the two peaks 902, 905 in the tracer gas signal are detected by delta detector 705, which is essentially a slope detector for finding peaks in the tracer gas signal. The tracer gas and EtOH signal samples, preferably recorded at a rate of six or more samples per second, undergo addition and subtraction operations in real time to detect upward or downward movement, thereby enabling the detection, time of occurrence, and quantification of peaks above a somewhat noisy background. The tracer signal features are peaks 902 and 905, surrounded by background values ​​901, 903, and 904, 906, respectively, acquired immediately before and after each peak. From these values, the peak magnitude can be calculated. The use of multiple background points allows for continuous monitoring of background fluctuations, which may be important for the selection of schemes according to Figures 7 and 8 (see below). The fluctuations in the background signal may be referred to as "noise." Such "noise" may be fundamental in nature and / or dependent on environmental factors such as temperature, humidity, transients, etc. Preferably, "noise" is measured in real time by continuous and automatic sampling of both the tracer gas signal and the EtOH signal between the periods of peak detection and peak quantification. For example, by comparing the root-mean-square (RMS) fluctuations of the signals, the noise level may be quantified and related to the calculated BrAC peak level.

[0071] A BrAC measurement will be derived from the tracer gas and EtOH signal peak values ​​using Equation (1). If this measurement is much higher (or lower) than the legal concentration limit, the BrAC measurement may be used directly to classify the concentration as "high" or "low," supported by comparison with the actual noise level. For "intermediate" concentrations, the actual noise level will similarly provide appropriate decision support. The BrAC measurement taken in conjunction with the legal limit and noise level for the particular jurisdiction in question will therefore determine the level of confidence with which to classify the breath alcohol concentration as being at a "high confidence level," "intermediate confidence level," or "low confidence level." By adding more BrAC measurements, the confidence level of the BrAC measurement increases (e.g., from "low confidence level" to "intermediate confidence level" or "high confidence level").

[0072] Combining data on systematic measurement error, the noise level may be used to define the confidence level of the BrAC measurement using equation (1). The timing of events 901-906 in the tracer gas signal may connect the events to corresponding points 1001-1006 in the EtOH signal, allowing for calculation of the magnitude of the corresponding EtOH peaks (i.e., EtOH peaks 1002 and 1005 correspond in time to tracer gas peaks 902 and 905, respectively, and EtOH background values ​​1001, 1003, 1004, and 1006 correspond in time to tracer gas background values ​​901, 903, 904, and 906, respectively; note also that the two peaks 1002 and 1005 in the EtOH signal are detected by delta detector 806, which is essentially a slope detector for finding peaks in the EtOH signal). From the magnitude of the tracer gas peaks in FIG. 9 (i.e., peaks at approximately 0.1), the DF may be estimated to be approximately 80, allowing for an estimation of BrAC = 0.25 mg / L according to Equation 1 above.

[0073] If the BrAC values ​​based on the first and second peaks differ such that classification among "high," "medium," and "low" concentrations with respect to the legal limit cannot be made within a specified confidence level (e.g., "high confidence level," "medium confidence level," or "low confidence level"), an active breath test may be required to be obtained from the driver, or an overruling principle, e.g., a principle that minimizes the risk of personal injury, may be used to determine vehicle drivability (e.g., suspending vehicle drivability until the BrAC is determined to be below the legal limit for that particular jurisdiction with a "high" level of confidence).

[0074] The procedure described above will automatically compensate for slow variations in background values ​​in both the tracer gas and EtOH channels. The use of continuous or digitized signal acquisition followed by signal accumulation provides an increase in the signal-to-noise ratio.

[0075] The system overview of Figure 8 may be advantageous when there is a stable background tracer gas concentration (which may be identifiable by relatively stable background values ​​901, 903 and 904, 906). In such cases, there will be a very strong correlation between the two signals (i.e., between the tracer gas signal and the EtOH signal). In the presence of occupants or other external sources of background variation, the scheme shown in Figure 7 may be more advantageous.

[0076] Essentially, the present invention comprises a novel method and apparatus for utilizing multiple passive breath tests to determine whether a driver's calculated BrAC concentrations (i.e., cumulative BrAC concentrations) should be classified as "high" or "low" overall. Additionally, the present invention allows for the determination of whether the cumulative BrAC concentrations provide an appropriate confidence level (e.g., "high confidence level," "medium confidence level," "low confidence level") for a particular classification of the BrAC concentration, with the confidence level being, in part, a function of the total number of breaths analyzed to arrive at the cumulative BrAC concentration.

[0077] Modifications of the Preferred Embodiment It will be understood that the foregoing description is merely illustrative of the principles of the invention, that the invention may be practiced in embodiments other than those described, which are presented for purposes of illustration and not limitation, and that the invention is limited only by the claims which follow.

Claims

[Claim 1] 1. A method for passive breath alcohol detection, comprising: A) passively obtaining a first air sample from the air within the interior of the vehicle; B) determining the concentrations of (i) the tracer gas and (ii) the analyte present in said first air sample; C) passively obtaining a second air sample from the air within the interior of the vehicle; D) determining the concentrations of (i) the tracer gas and (ii) the analyte present in the second air sample; E) passively obtaining N air samples from the air within the interior of the vehicle, and for each air sample obtained, determining the concentrations of the tracer gas and the analyte present in the air sample; F) determining the number of tracer gas concentration peaks and the number of analyte concentration peaks present in each of the air samples; G) determining a confidence interval based on the number of tracer gas concentration peaks and the number of analyte concentration peaks; H) controlling operation of the vehicle based on the confidence interval and a function of the concentration of the analyte present in the air sample.

Images