Motivations
- Beekeeping as a craft/hobby seems objectively fun and rewarding
- Low overlap of tech-savvy people with population of beekeepers means much opportunity for invention and improvement
- As a non-zero-sum activity, the community tends to be welcoming, helpful and positive
- Personal interest in bees from a sci-fi and machine learning perspective, since they perform distributed tasks well and display emergent properties of intelligence
- Honey is tasty
Sources
My starting research sources included:
- Doolittle's "A Year' Work in An Out-Apiary" from 1907
- Beesource Forums
- Howard County Beekeepers Association
- NASA's hive weight project
- NASA's honey bee forage species for my region
Initial Decisions
Some of the early decisions I made when selecting equipment:
- What size hive boxes should I use? I'm a pretty strong, young guy so I went with "all-deeps" for all of my hive boxes and frames. This means all the frames are the same size, and any frame can go anywhere. We'll see how I feel about this next year.
- Starting bees: I will pickup a nuc in late April from Honeysmith, a local bee company. I will also try to catch a feral swarm.
- How will I keep my hives from swarming? Checkerboarding. Hopefully.
- What is my goal number of hives: 3 steady-state hives, no more than 4. I don't have space for many more in the yard than that.
Open Questions
I have many questions about the specifics of beekeeping in my region, a factor that includes the weather conditions around my house and the relevant plants and water sources in about a 7km radius around the hives. Most of these will help me answer the question of when should I rearrange the hive or add more frames in order to keep the hive from swarming and maximize honey production?
Data-driven Approach
I was inspired by NASA's hive weight project to try collecting some data about my own hives to inform my answer to the above question. With hive weight data, I will be able to estimate when the hive is getting full and needs more frames. I will also be able to observe issues like robbing or low internal temperatures, hopefully earlier than I would without the data. Finally, I will be able to use historic data from previous years to predict when honey flow will happen.
I could gather the same data manually, but that requires a lot of repeated visits to the hive and possibly disturbing the hive. Instrumenting the hive is the obvious solution.
What data will I gather?
I plan to collect at least daily, but hopefully more often:
- Hive weight
- Hive internal temperature
- Hive external temperature
I can use already gathered data for correlations:
- Weather temperature
- Humidity
- Rainfall
- Sunlight hours
- Air pollen count
I might add on, if I feel this data would be helpful:
- Audio recording with some processing to extract principal frequencies. Reportedly, non-queenright hives will make a different humming noise than queenright ones.
- Infrared gate counter at the hive exit, to track traffic