Safety managers and quality or technical staff who review breath alcohol test results cannot judge a result's reliability from a single concentration reading on the screen. They need to look separately at whether the exhaled breath was accepted as suitable for testing, how the device processed the signal from the sensor into a result, and whether that result carried the same meaning when moved to the screen, a smartphone, or a PC. This is why a difference in numbers alone is not enough to conclude that there is a sensor fault or a measurement error.
An electrochemical sensor is a sensor that generates an electrical signal in response to alcohol in the breath. DATECH's official product information for the AF-40BT states that the product uses an electrochemical sensor and supports storing test results and connecting to a smartphone over Bluetooth 5.0 LE. However, the sensor type is only the starting point of the conversion path. To trace the path from sensor output to a concentration reading, you need to separate breath sampling, signal analysis, display, and storage and transfer.
Separate the path that produces the number
A test result usually passes through several stages: receiving the breath, obtaining the sensor signal, analyzing the signal and processing it into a concentration reading, and displaying it on screen or saving it as a record. If the result is moved to an external app or PC, a transfer stage is added. Knowing that the sensor is electrochemical does not tell you the settings and processing used at each of these stages.
In particular, the number on the screen is the final output. When two screen readings differ, you cannot tell from the screen alone whether the breath input differed, whether the device software or applied settings differed, or whether the unit and decimal display differed. Conditions that should not be compared as concentration readings, such as insufficient breath or a test that did not complete normally, must also be viewed separately from numeric results.
| Check segment | What to examine | Why it must be separated |
|---|---|---|
| Breath sampling | Test start and end times, validity judgment of the input | To avoid reading insufficient breath as a difference in concentration |
| Signal analysis | Time window in which the signal was received, device software version, applied settings | It is the basis for judging whether results came from the same conditions |
| Display, storage, and transfer | Display unit and digits, normal completion, transfer time | You can find where a difference arose between the screen value and the stored or synced value |
The items in the table do not mean that every device provides them in the same form. Rather than making assumptions about unpublished internal signals or calculation coefficients, it is better to start by collecting the items you can actually check on the current device and its connected program.
Do not separate units and digits from the value
AF-40BT official information lists the display range in %BAC, ‰, and mg/L. Accuracy is also given together with each condition: 0.100 %BAC, 1.00 ‰, 0.50 mg/L, and 25 ℃. So when you store or compare results, do not record only the number. Also record the display unit and test conditions at the time.
Numbers in different units may be different expressions of the same test result, so they must not be compared side by side as if they were separate readings. The number of decimal places on screen is also subject to review. A photo of the screen shows what was displayed at the time, but if the stored record lacks the unit, test time, or normal completion status, the photo alone cannot be used to judge a problem at the sensor stage.
You can trace the path without knowing the internal formula
A manufacturer may not disclose all of the raw electrical signals, correction coefficients, and analysis algorithms. In that case, if the goal of the review is to reproduce the internal formula, the scope you can check becomes too narrow. A more practical approach is to check what the device and the connected system each display and store during a normal test, an invalid breath input, an interrupted test, and a communication retry.
As a hypothetical example, imagine that a photo of the device screen shows a result, but the smartphone app record is shown in a different unit or is missing. First, look at the test time, unit, and normal completion status in both the screen photo and the app record. If the time matches but the unit or completion status differs, you can first interpret it as a difference in display or transfer processing. If there is no app record, check separately whether the connection and transfer were completed at that time.
One possible action is to match the result left on the device with the record the app received, using the test time and any record identification the device provides. If the link between the original result and the synced record can be explained, and the cause of the unit or status difference or the missing record is confirmed, the trace for this case can be closed. On the other hand, if there is no evidence of a link, it is right to leave it as an unconfirmed part of the sync path rather than conclude that the sensor is faulty.
In comparison tests, fix the point of change first
When you compare an upgraded device with a previous one, or environments with different app versions, do not compare only the numbers on the screen. First write down what changed among the sensor, device software, user settings, and connected app. Then, under the same input conditions where possible, see at which stage a difference appears. This order keeps you from mistaking a change in display format for a change in sensor performance.
For example, if the reading on the device itself is the same but only the number of decimal places in the exported record has changed, that fact alone is no basis for saying the sensor has changed. On the other hand, if one result did not complete normally, you must first find the reason for the status difference before comparing numbers. In the comparison record, it is best to note, within the confirmed scope, which segment the difference belongs to: the sensor, analysis processing, display format, or transfer processing.
In storage and sync, you must be able to find the original result
The AF-40BT is described as storing up to 10,000 test results and supporting smartphone connection. However, having storage capacity and wireless connectivity does not mean that all the information needed for review is kept automatically. You need to check separately what the device stores and what the app or PC receives and keeps.
Operating records can be set up so that the reading, unit, test time, normal completion status, and, where the device provides it, information that identifies that result are kept together. When a problem occurs, match in order which device generated the result and when it was moved to the external system. By dividing the conversion path this way, you can tell a sensor signal problem from a display or transfer processing problem without guessing about undisclosed internal circuits.
