Team HiveMIND

We’re team HiveMIND, a sophomore Gemstone team studying the use of neuromorphic hardware and event based vision to monitor beehive health. Gemstone is a UMD Honors research program in which teams of undergraduate students propose research projects, perform experiments, and present their original research findings. Our team consists of 11 students who are passionate about using computer science and engineering principles to learn more about how we can protect the environment. Specifically, our team plans to use an innovative vision sensor known as a dynamic vision sensor to monitor bees.

The vast majority of the world’s tropical forests are pollinated by bees, and even in temperate climates smaller plants like shrubs and wildflowers depend on bees for reproduction. Of particular interest to humans are agricultural crops such as fruits, vegetables, fiber, and forage. Most top monocultures in agriculture require pollination by bees, and in the United States bee-pollinated crops are worth an estimated $18.9 billion. Declines in bee populations have the potential to have sizable impacts on the biodiversity of flora, especially in regions where pollinator services are highly specialized. The health of many ecosystems globally would be devastated by such a change, highlighting the impact bees have on their overall health.

The dynamic vision sensor is modeled off of biological eyes, and responds to changes in the environment rather than absolute brightness levels like a traditional camera. As a result, it is extremely useful for tracking small, fast objects like bees. Our team plans to use the DVS to perform operations such as counting bees to track population levels, as well as monitoring the hive for signs of invaders in real time.