Back to projects

Buzzing into the Future with Computer Vision

Tracking bee populations one snap at a time.
Productization
Machine Learning
React Native
TensorFlow
IGNITING BEE TECH:

Turn insect enthusiasts into field researchers

Some people see a bee and run away. Others grab a camera! Just like birders, there’s a virtual army of bee-spotters out there, ready to identify and catalog all the species of bees they can find.

Brian Spiesman, Entomologist and head of the Pollinator Lab at Kansas State University, developed a promising AI model that could identify the genus and species of bumblebees just from a photo. Powerful technology, ready for prime time.

From Hive to Screen: Digitizing
Bee Discoveries

Spiesman partnered with Germinate to put the model into a native mobile app that anyone could use. BeeMachine was born, and published to the app store. Wrapped in an easy and fun user interface and available widely in app stores, insect enthusiasts everywhere could become backyard scientists and field researchers.

And whats more, the effective reach of the pollinator lab became limitless. With an easy to use app capturing and identifying bee pics right on your phone, the Pollinator Lab starts collecting meaningful data about bee population around the globe.

Scientists win, insect lovers everywhere win, and (from what we know about the importance of pollinators) the whole world wins!

Buzzing into Efficiency: How Germinate collaborated with Advanced Solutions.

The Challenge

1. The AI model was powerful but not easy to use
2. Each team held different expertise in the project
3. The interface needed to make a complex subject= simple and fun

Design Principles & Insights

1. Use existing standards that app users will be familiar with
2. Start small and add features over time
3. Create a development pipeline that allows contribution from different organizations

The Solution

1. Firebase for a quick and scalable app back-end
2. Figma for rapid prototyping and design systems for smooth integration
3. TensorFlow integration for quick, on-device AI processing (Object Detection + Image Classification)
4. Expo open-source platform

Leading the Swarm: Guiding the team through the development journey.

On-the-Go Bee
Sighting for quick
Data Collection
Product screenshot

Bee sighting that allows you to take a picture of your bee sighting, the date and place where you find it and notes. You can even see the % of species that that bee could be.

"Germinate was able to take our initial AI models and prototype app and turn it into a fully functional, easy to use mobile app that anyone in the world can use. Their partner-oriented approach allowed us to stay involved in the technology along the way, so everyone was able to contribute and move things forward”

Brian Speisman, Pollinator Lab,
Kansas State University

Other Case Studies

Magma

Revolutionizing Investor Onboarding

Back to projects