Vertical farming is the practice of producing food in vertically stacked layers, typically in old industrial buildings in urban centers. For a number of reasons, some listed below, we see vertical farming as the future of agriculture. We're taking the first steps in creating GE's next big software offering - the GE Digital Farm.
- Market for locally grown crops grew from $1 billion in 2005 to $9 billion in 2015
- By the year 2050, nearly 80% of the earth’s population will reside in urban centers
- The human population will increase by about 3 billion people during the interim
- An estimated 109 hectacres of new land (about 20% more land than is represented by the country of Brazil) will be needed to grow enough food to feed them
- We're already using 80% of suitable farming land to grow crops
Environmental Impact Water
- 70% of our water supply goes to agriculture
- 70% of water contamination comes from agriculture
- Vertical farming uses 95% less water
- 25% of the world’s land is now “highly degraded,” with soil erosion, water degradation and biodiversity loss
- Vertical farming uses less than 1% of the land required by conventional growing
- Vertical farming is over 75 times more productive per square foot vs. traditional agriculture
- 95% of America's leafy greens are grown in the western portion of the US, meaning it may have to travel thousands of miles to get to your plate.
- Vertical farms can exist in old industrial buildings within any city, meaning transportion costs (financial and environmental) are mostly eliminated.
What it does
Provides vertical farms with a way to visualize their KPIs, pull insights from them, and adjust their operations accordingly.
How we built it
For a production build, this will require many more sensors (type & quantity) - we used the provided Intel Edison board and Predix Machine to simulate some of the sensor data that will be used in production, light and temperature. We pulled this into the Predix Time Series service from which we used custom algorithms to feed it into Predix UI in insightful ways. Using existing Predix components, we were able to create charts to provide visualizations to help vertical farm owners gain quick insights.
Challenges we ran into
- Finding data - ideal "light recipes" for LEDs is considered proprietary information (against the rules for this hackathon)
- Predix-UI context browser is trickier than expected
- Writing algorithms is hard
Accomplishments that we're proud of
- Successfully ingesting data into Time Series via Predix Machine
- Custom algorithms for Vertical Farm specific KPIs
- Chart manipulations (trickier than we thought)
What we learned
- How to use Time Series
- How to use Predix Machine
What's next for Digital Farm
- Work with Current, powered by GE to create a full build and begin selling to customers.