Telling overlapping timestamps apart after a breathalyzer clock is set back

6 min read

When you store breath alcohol test results on the device or move them to a smartphone for management, the date and time are the basis for finding and sorting records. A timestamp is that date and time attached to a result. But if the breathalyzer's internal clock is set back to an earlier time and a new test is taken, the earlier stored result and the later result can end up with the same timestamp. If you delete one of them as a duplicate just because the same time appears twice, you can lose an actual test result.

This problem is not about whether a reading is high or low; it is about determining which single test each result came from. DATECH's official product information for the AF-40BT describes storage of up to 10,000 test results and Bluetooth smartphone pairing. When reviewing stored results in an app or another system, use the time as a sorting clue, but do not treat the time itself as if it were a unique number for each test.

The same time and the same test are different things

After the clock is set back, a 10:15 a.m. record created earlier and a 10:15 a.m. record created again after the clock change can both remain. Even if the readings are the same, you cannot conclude that the two records represent the same test of the same person. Conversely, you cannot confirm that the clock was changed just because the values differ.

So the review starts by separating the observation "the times overlap" from the judgment "the same result came in twice." The time is information that narrows down the candidates to compare; whether something is a duplicate can be judged only by looking together at the identifying information the device and integration software actually provide, the number of stored results, and the transmission path.

First separate original test records from re-received rows

Overlapping times are not always caused by a clock change. If a smartphone app or PC program fetches previously received history again, one test can be stored as two rows in the external system. In that case, the test was not taken twice; the same original result was received twice.

So it is better not to compare the results stored on the breathalyzer and the rows newly created in the app or PC as if they were the same thing. If there is one such result on the device but two in the external system only, look into the possibility of re-reception. Conversely, if two results with the same time are confirmed in the original list exported from the device, you must not remove one of them using only the external system's deduplication function.

Set grounds for comparison beyond the time

Use only the information the device and integration method actually provide as comparison criteria. The official AF-40BT product information mentions a cumulative test count, but that alone does not mean each stored result is assigned a unique number. You must not separate two records by assuming boot numbers, communication identifiers, or reset history whose availability has not been confirmed.

Information to compare What to check How to use it in the judgment
Breathalyzer and model information Whether the results came from the same device If the devices differ, keep them as separate results even at the same time.
Result number shown by the device or app Whether there is a number that distinguishes individual results Distinguish whether the number is a value from the device original or a management number the app assigned.
Export order and time received When and in which batch it was transmitted Use the reception order only as a supporting clue and do not equate it with test order.
Confirmed material on clock changes Over what range the time conflict occurred Divide analysis groups only within the range where the change is confirmed.

If the app assigned new row numbers during import, those numbers can be used to tell two rows apart within the app. However, they do not prove two different tests inside the breathalyzer. Likewise, the row that appears first in a file may represent transmission order rather than test order when past history was sent all at once.

Record observations and conclusions separately in a hypothetical case

For instance, suppose that after moving a breathalyzer's results to the app, two rows are found with the time 2024-05-10 09:30 and the same reading. The first row is in data received in the morning, and the second is in data synced again in the afternoon. The observation you can record at this stage goes only as far as the fact that the same time and value appear in two files and that they were received at different times.

The interpretation starts from the point that you cannot yet determine whether the second row is a new test or a re-reception of the first original. If the result number provided on the device or integration screen is the same, and only one stored result can be confirmed on the device, there is a basis to treat it as a re-reception. Conversely, if there are two results with the same time in the original list, or the provided result numbers differ, keep them as separate tests even though the time and value are the same.

A possible action is to record the judgment and its basis next to each row. If you treated it as a re-reception, mark which result it was linked to; if it is a separate test, record that there was a time conflict. The completion criterion is not reducing the number of rows but reaching a state where you can later recheck whether each row is an original test, a re-received copy, or undeterminable, and why.

Keep both records when they cannot be told apart

If you could not obtain any of the result number, the device's stored list, or the reception history, you cannot determine the relationship between the two rows. In that case, it is appropriate not to delete one as a duplicate but to mark them as "same-time identification conflict, relationship unconfirmed" and keep both. If you remove confirmed duplicates and possible duplicates by the same rule, it becomes hard to explain later which result disappeared and why when additional material turns up.

You also must not infer the exact time of a clock change, a device restart, or a storage reset just from the fact that times appear in reverse order. Only when there is material showing a clock change can you divide the before-and-after range it points to for analysis purposes. It is good to make clear that any segment names assigned here are labels for review convenience, not values output by the breathalyzer.

In future integrations, do not depend on time alone

For results coming in from now on, you can organize them in a structure that keeps comparison material other than time. If you keep the values actually available, such as individual result information provided by the device, breathalyzer information, and which batch of data you received, you have grounds for judgment even when times overlap. Even when using a Bluetooth connection, one connection does not always correspond to one test, so do not interpret connection records as test records themselves.

The key is not to arbitrarily correct the original times so they sort neatly. You need to keep the time shown in the original and explain the relationship between results using the provided identifying information and the storage and transmission path. Leaving records without enough grounds for explanation as unconfirmed reduces the chance of wrongly removing a different test as a re-sent copy.

References

#AF-40BT
#timestamp conflicts
#clock change
#duplicate records
Considering our products?
We'll help you decide.