How NASA Is Using AI to Help Hawaii Become More Food Self-Sufficient

The state of Hawaii imports up to 90 percent of its food, leaving its residents at an increased risk of food insecurity due to supply chain disruptions and high prices. Now a NASA project aims to take on this issue by employing AI to help grow more food locally, starting with a trial project on the island of Maui, according to Food Dive. 

The initiative relies on building an agricultural dashboard to track detailed crop data using satellites and AI, according to Hannah Kerner, an assistant professor at Arizona State University's School of Computing and Augmented Intelligence.

By gathering on-the-ground information from farms across Maui County, the platform will allow officials to monitor crop health, gauge yields and respond rapidly to shortfalls. The predictive models even hope to fill gaps left by intermittent government agriculture surveys. 

Developed under the NASA Acres and Harvest consortiums, which are focused on using satellites to improve food security and ecological resiliency, the dashboard brings more stability to cultivation through data-driven insights.

To feed the predictive models, Kerner, who is the AI and Machine Learning Lead at Harvest, collaborated with the University of Hawaii Maui College to gather precise field data. This involved extensive surveys of local farms and interviews with farmers to collect accurate, real-world information about land usage and cultivation issues.

By combining this ground-truth data with satellite imagery analysis, the dashboard could forecast shortfalls in food production across the island and illuminate gaps in access through machine learning.

The project has to contend with the state’s specific farming challenges. Hawaii is home to over 7,300 modestly sized farms often situated on the islands' jagged inclines and rough terrain, making large machinery impractical. Laborers toiling with hand tools frequently replace the tractors and combines prevalent elsewhere, driving costs up and reducing productivity compared with more highly automated operations. 

The dashboard could help plan around these issues and guide long-term agricultural planning and adaptation. By analyzing NASA data over time, regions suitable for farm expansion could be identified. The system may illuminate why crops are struggling in certain areas and which alternate crops could thrive there instead.

It could even help determine the impact of invasive species like guinea grass on fallow farms — a contributing factor to the August 2023 Maui wildfires' severe damage. With richer data, farmers could pinpoint the most cost-effective cultivation methods for their land.

The first crop cultivation map created by Kerner's team tracks the growth of taro — a nutritious, sacred crop in Hawaiian culture. 

But the dashboard has applications beyond agriculture. After the August 2023 wildfires, NASA rapidly mapped burn zones to assist Maui United Way's distribution of $7.5 million in disaster funds to Lahaina residents within two months.