When breathalyzers are placed at several branches or entrances, each device can produce its own test result files. A breathalyzer is a device that analyzes breath and displays or stores alcohol-related test results, and some sites use the test results in the next step, such as access control, after checking them. Buying several devices means not only adding test locations but also deciding on what basis to bring the scattered results together.
What the purchasing staff need depends on whether the requirement is simply gathering files into one folder, regularly totaling results by branch, or quickly checking results from several devices on one screen. If records from several devices must be combined, do not compare only the number of devices and people tested. Include in the purchase conditions the information that identifies devices, how the same file is handled when it comes in again, and how missing data is found.
The first thing to decide is when integration is needed
Integration does not always mean a setup in which a central system shows every device in real time. Depending on the operating purpose, it may be enough for each site to export test result files on set days and for the person in charge to gather them on one PC. On the other hand, if results must be confirmed immediately at the time of entry, gathering files later will hardly meet the requirement.
So when requesting a quote, it is better to state when data will be gathered rather than writing something broad like "whether records can be integrated." For example, distinguish whether each branch's files will be gathered after closing every day, whether head office will tally them weekly, or whether results must be reflected in one place right after testing. Only when the needed timing is clear can you compare whether the setup should center on PC connectivity or needs a separate integrated screen.
Distinguish device names from actual devices
When combining results from several devices, the first thing to check is whether each record shows which device took the test. A nickname given at the site, such as "Seoul warehouse No. 1," is easy for people to recognize, but if the device is moved to another branch or the nickname is changed, past and new results can be confused. This is why you must check how the device name people see is linked to the identification information that distinguishes the actual unit.
Especially when buying several units of the same model, the model name alone cannot tell you where a record came from. It is best to ask the supplier to show whether results can be viewed separately by device when data from several units of the same model is loaded together, and whether the previous device's results disappear when a device is replaced. At sites where devices are moved or replaced often, the history of nickname changes is also something to check.
| Item to decide before purchase | What to see from the supplier | Reason for the decision |
|---|---|---|
| Identification by device | How results from several units of the same model are each identified | To trace which branch a result came from |
| When data is gathered | Possible gathering methods, such as right after testing, daily, or weekly | To choose a setup that matches the speed of confirmation needed |
| Re-entry of the same data | What happens when a file already processed is loaded again | To prevent totals from increasing unnecessarily |
| Checking for missing data | How it is shown when data from a particular device or branch is missing | To find data not submitted or that failed to transmit |
Duplicates are not something to delete automatically
When gathering several files, the same test result may come in twice. However, a result from the same person testing again within a short time must be distinguished from a result created by reloading a file already imported. The former may be a test actually performed twice, while the latter may be reprocessing of the same data.
Take a hypothetical example. Devices at Branch A and Branch B each stored test results at the same time, and the next day the person in charge gathered both files on one PC. If, after the Branch A file alone is imported again, only the total count increases, the person in charge will find it hard to tell whether it was an actual retest at Branch A or a file re-entry. What to observe in this case is the count per device and the device identification in each result rather than the total, and the interpretation depends on how the system judges identical files.
A possible action is to request, in a pre-purchase demonstration, a test that imports the same file twice and a test that loads data from different devices at the same time together. The completion criterion is a state in which results from two genuinely separate devices are kept, while the person in charge can see how a re-entered file was handled. Do not assume a particular behavior, such as automatic deletion or a separate mark, without confirming how the product and program actually behave.
Inspect with data from several devices, not a one-device demonstration
The fact that PC software is provided does not mean every device from the same manufacturer connects the same way or supports the same scope of data handling. DATECH's official Downloads section posts information on downloading the ALCOFIND PC software, so it is safer to confirm specifically at the quote stage the combination of candidate models and the software to be used. If you already have devices, you should also give their model names.
Rather than ending the demonstration with one device reading one file, it is more suitable to put hypothetical data from different devices into one environment. In doing so, compare the number of results per device with the number after integration, and deliberately leave out some branch data to see whether the omission is noticed. Also checking whether, when an error occurs during processing, only that data can be reprocessed while keeping the original files helps you prepare for situations where data is hard to gather again, such as month-end tallies.
Put conditions other than the number of devices in the quote
It is best for the quote request to include not only the installation locations and number of devices, but also the model names of devices currently in use, how often data will be gathered, and who will view results. For example, if head office staff should see results from all branches while branch staff should see only their own branch's results, you need to confirm whether the setup allows that viewing scope. Gathering results centrally is no reason for every staff member to see everyone's test results.
It is better to record the purchase conditions as specific test items rather than a single phrase such as "multi-device integration." Compare, for each candidate setup, whether you need identification of several units of the same model, handling of file re-entry, checks for missing data, and viewing of past results after a device replacement. Once the needed integration timing and inspection criteria are set out, you can choose a combination of devices and software that avoids an excessive central setup or, conversely, a simple file-gathering method that does not fit the work.