The Bee Informed Partnership’s Electronic Hive Monitoring program gathers data from hundreds of electronic hive monitors located all over the USA. Much of this data is publicly available to the beekeeping community in a variety of actionable formats. For example, if you go over to the BIP research portal at research.beeinformed.org and click on Hive Monitors you can see a public gain/loss map that shows the 7 day moving average weight loss/gain of the hive monitors within that state.
We are also doing algorithmic work on the data and attempting to discover more actionable information about colony health from this vast collection of electronically captured data. If you are a participant in BIP’s hive monitoring program, we would like to share some advise with you on how you can make your data more useful to the research community. One of the most important things we need you to do is to login to the research portal using the same credentials that you registered your scale with. Once logged in we need you to actively interesting changes in the data via our annotation interface. For example, if you’ve added an empty honey super to the hive, you’ll likely see a 20 lb increase in weight on that day. We need you to find those types of changes in weight and tell us what happened to cause it. In the future this additional info will be very helpful in helping us improve our machine learning algorithms which we hope will ultimately help the beekeeping community make better colony management decisions.
To help simply this task for you, we’ve already introduced a feature to our research portal to automatically find all the changes in weight that likely require some sort of human annotation. Every night when you’re getting your rest after a long day of beekeeping activities, our algorithms kick in and analyze all the data received over the past 24 hours. Any changes in weight that likely need annotation will be flagged and reported to you the next time you login to our portal. To get a better idea of this works, take a look at the YouTube demonstration video below.
We hope this short demonstration of how easy it is for you to annotate your data has motivating you to login to the BIP research portal and get busy marking up your data! Remember, the more annotations you provide on your data, the more useful your data becomes as we continue to try to build even better machine learning algorithms!