Data Analysis and Validation in Temperature Mapping
What Do You Do With All That Data?
After the mapping study is complete, when you’ve collected all of your labeled EDLM’s from your temperature mapping locations, and have them in a container with pages of their associated data sheets at your desk, it’s time to start the next step of the temperature mapping process: analyzing your data. Below are the steps you should take after you have all your mapping EDLM’s out of your facility (or you have all of their data, delivered wirelessly at your desk).
You need to download your data! If you used a wireless data monitoring system, this could be as easy as going into your cloud-account and clicking download, or visiting your data file folder and dragging the information to your desktop. If the data loggers you used in your mapping study connect to your computer via USB, plug them in and start downloading!
Immediately upon downloading your temperature mapping data, back it up. We can’t stress this enough. Save it to a Dropbox folder, save it to shared company drives, save it to Google Drive, whatever. Just make sure that your temperature data is not just stored on your local hard drive. It should be saved to multiple locations, for disaster prevention.
3. Find Failures
The next step is to find those sensors that failed to record data. If you are mapping a large facility, at least a few sensors will fail: it’s just the nature of temperature mapping. Whether they fail the entire temperature mapping study, or just a portion of it is for you to find out. If you find a failed sensor, there is an accepted process for accounting for that failed location. You should take the averages of the 3 closest sensors to the failed sensor location, and average out their data for the time range that the sensor failed for. As long as you weren’t being too sparse with your sensor placement, the averages of those data loggers will suffice. Just be sure to document it!
We wish we could say outright that this step was completely automated for you (if you are using a contractor to map your facility, it will be) but it’s not. You should know what averages and calculations you will need to make from the EDLM’s temperature and humidity data before you start the study. Now is the time to calculate it. Some EDLM software can calculate values like Mean Kinetic Temperature for you, and we highly advise using such software (like the new DicksonWare) to calculate your averages and means.
5. Find Deviations
Next up is the scavenger hunt. Ideally, your storage area has a single, perfectly uniform temperature. But, that won’t be the case. This next step may be the most important. What you or your contractor will first look for, is temperatures that are out of the acceptable range. This can be done via a graph-overlay. Ideally, you want to take the data from each of the EDLM’s and place their graphed data over top of each other. You should either create physical or mental reference lines within your EDLM software of the highest and lowest temperatures that your products can be stored in. If at any time during the study an EDLM recorded temperature data outside of that acceptable range, you know you have a problem spot.
That may seem simple enough, but finding deviations is a little more nuanced than that. You should also look for trends of sensor groups towards extremes. If you notice that sensors placed in the southeast corner of your facility read temperatures that were 2-3F colder at night then sensors in the rest of your facility, you know that you have a potential problem spot in that corner.
Once you’ve found deviations in your facility’s layout, it’s time to investigate. This step should be done in haste, especially if a product is located where the deviation occurred. First, you should remove product from the deviation locations until the EDLM temperature reading can be corroborated with another sensor. Next, you should corroborate the EDLM reading with a continuous monitoring device. Checking that device frequently in the time after the mapping is crucial to the investigation process, as you will want to not only corroborate the temperature excursion, but the timestamp for that temperature excursion. Finding the cause and then fixing the problem spot is the final step of your investigation.