Hopefully some of you have already seen one or two of the articles related to the “bee project” that has been taking place over the past few months and has come to an end (at least for now). For those of you who have missed the posts or not seen them all here is a round-up article giving you a brief overview of all the fellows involved, the tasks they had, the progress and the current state of the project.
You may already be aware of the importance of bees to our planet, but for those who aren’t here are a few facts. Bees are responsible for pollinating essential crops, like fruit and vegetables but also tea and coffee. The majority of crops rely on bees for pollination to increase the amount and quality of the produce. By supporting the growth of various plant species, bees also allow a diverse number of animal species to coexist. This makes bees and essential part of our ecosystem.
Unfortunately, the bee population is declining with some – once common – species becoming endangered. To combat this loss, there has been a rise in beekeeping, and SAP has joined the movement. With four bee hives in St. Leon Rot, one in Markdorf, and the latest one on the roof of the SAP AppHaus in Heidelberg, the aim is not only to keep bees but also see how SAP can apply its expertise to help bees thrive by analyzing data that comes from the hives. Here’s more about the bee project.
How it all started
To explain how the project started we have to go a few years back, to when our colleague Michael Koegel mentored Johannes Weber, the founder of StadtBienen e.V., as part of the SAP Social Impact program (now called starter). StadtBienen produces ready-made bee hives called “Bienenbox”. Michael was so intrigued that he started as a hobby bee keeper himself and initiated the setup of the bee hives at SAP.
The project we want to tell you about today began when the start-up company “BeeandMe” got in touch for a joint project in 2017. Michael referred them to the d-shop where the idea formed for an endeavor that would bring SAP and the start-up together while showcasing SAP Leonardo. The aim is to show how modern technology like the beehive scale (developed by “BeeandMe”) and machine learning can improve the lives of bees, the work of the beekeepers and provide insights that so far can’t be generated today. A fellowship within the d-shop was advertised and so a team of six formed to work on this project. The topics the fellows worked on range from UX design (finding out what exactly the beekeepers could gain), to developing a bee counter (to determine how many bees come and go throughout the day) or analyzing the sounds generated by the beehive with machine learning.
The swarming phenomenon
The main problem beekeepers seem to have is “swarming”. This is a phenomenon which occurs when the beehive becomes so large, that is must be divided – with the old queen bee flying off with half of the worker bees to look for a new hive while the new queen bee takes care of the old hive. This may not seem to be a problem, but especially in urban settings, finding a new hive can be quite problematic and can lead to the old queen and her workers dying.
By monitoring the hive with the beehive scale and sound inspection it could be possible to predict swarming before it occurs, giving the beekeeper enough time to figure out what to do (i.e. find a new hive or prevent swarming etc.). Remote inspection has another advantage: it doesn’t disturb the bees like normal checks do and is thus less stressful for them.
Current state of the project
So – what has happened so far? What are the key take-aways from this project?
There have been a few set-backs with the technology provided – the team had anticipated a sound monitor to go along with the beehive scale. Correctly analyzing and interpreting the sounds coming from the hive, could help predict swarming. There are mainly two sounds which can be heard in the swarming period: The sound the new queen makes when she is ready so hatch, so called “tooting”, and the sound the old queen makes in response to this, so-called “quaking”. This is how the two queens communicate with each other. By analyzing sounds derived from YouTube videos, it was possible for the team to at least identify these two sounds even if it has not yet been possible to evaluate sounds from the hive in Heidelberg.
The planned bee counter however has been successfully installed and is transmitting data regarding how many bees leave and enter the hive and comparing this to temperature, wind and precipitation taken from the local weather station. Monitoring when the bees enter and leave the hive in relation to outside factors helps generate insights into their flight patterns, when they start gathering honey, and possibly determining a correlation between their activities and swarming.
Colleagues from St. Ingbert got in touch to potentially continue the project in their location, seeing as it gained a lot of interest when it was first published. So hopefully this is not the last we have heard from SAP Leonardo helping the bees.
Follow us on twitter (link to: https://twitter.com/SAP_designs)