AgriSense

Where precision meets sustainability.

What do we offer?

Services We Provide

Crop Monitoring

Our technology enables precise crop monitoring for informed decision-making.

KNOW MORE

Crop Monitoring

By leveraging advanced satellite imagery, we provide real-time data on plant health, growth patterns, and potential stress factors. This innovative approach allows farmers to make informed decisions, resulting in better resource allocation and improved crop performance.

Yield Prediction

We use historical data and current crop health to accurately predict seasonal yields.

KNOW MORE

Yield Prediction

By analyzing key indicators such as plant vigor and environmental conditions, we enable precise forecasting. This cutting-edge technology empowers farmers to anticipate yields with greater accuracy, reducing waste and increasing profitability.

Sustainable Farming

We promote sustainable farming methods for a healthier planet and thriving communities.

KNOW MORE

Sustainable Farming

Sustainable farming practices are vital for preserving our environment and ensuring long-term agricultural viability. Our platform monitors key factors like soil health and water usage, helping farmers adopt more sustainable practices while maintaining high productivity.

What we look for

Key Indicators

Soil Moisture

Soil Moisture[7]

We promote efficient irrigation practices by monitoring soil moisture to conserve water.

Chlorophyll Content

Chlorophyll Content[8]

Monitoring chlorophyll enables customized nutrient applications, ensuring optimal nutrition.

Surface Temperature

Surface Temperature

We utilize surface temperature data to assess crop stress and enhance agricultural productivity.

Vegetation Indices

Vegetation Indices

Assessing vegetation indices like EVI helps in monitoring crop health and optimizing resource use.

Learn & Share

Our Product in Action

Explore our product demo and experience how easy it is to get started.

Video
Behind the Scenes

Our Methodology

We start by automatically downloading satellite data from USGS ESPA and crop data from the NASS geospatial data gateway using custom scripts. These scripts ensure smooth, automated retrieval and processing via Amazon SageMaker. The data is then transformed to maintain consistent spatial resolution and coordinate systems, focusing on specific areas of interest.

Ground truth data from the California Strawberry Commission, detailing crop yields and acreage, is cleaned and structured for analysis. This data is aligned with the satellite data to provide comprehensive insights into crop conditions and yields.

Both the processed satellite data and the cleaned ground data are securely stored in Amazon S3, making them readily accessible for further analysis.

Data Pipeline
  1. Our model helps farmers assess field health by predicting crop yields using historical yield data and satellite imagery, focusing on strawberry farms in Monterey County. Below are the details on how it is structured:

  2. Yield Data (Monterey County):
    • Temporal Mask: Includes only data from months when strawberries are grown.
    • Labels: Historical and current yield data (pounds per acre).
  3. Satellite Data (Monterey County):
    • EVI (Enhanced Vegetation Index):
    • Temporal Mask: Includes only data from months when strawberries are grown.
    • Geospatial Mask: Includes only strawberry farm data.

  4. Feature Engineering: Time Features capture seasonality using cyclical encoding for month/day-of-year and cumulative volume.
  5. Model Training: We employ a CNN Feature Extractor, which identifies regions of high and low productivity from satellite imagery.
  6. Hybrid Model: Combines CNN and LSTM for yield prediction.
  7. Prediction: Predicts yield (pounds per acre) for each pixel in user-defined regions.

Model

Compute
We use Amazon SageMaker to build, train, and deploy machine learning models efficiently. SageMaker ensures robust security throughout the process.

Storage / Database
Amazon S3 provides secure, scalable storage for all our data, ensuring durability and easy integration with other AWS services.

Front End
Using Streamlit, we create interactive dashboards that connect to our machine learning models for real-time predictions and analysis. This setup benefits from AWS's security features, protecting both the application and underlying data.

Infrastructure

  1. [1] “Water Scarcity”. World Wildlife Fund, 11 Nov. 2014, Link.
  2. [2] Khokhar, T. “Globally, 70% of Freshwater is Used for Agriculture”. World Bank, 22 Mar. 2017, Link.
  3. [3] Crosby, G. “The Nutrient Challenge of Sustainable Fertilizer Management”. U.S. Department of Agriculture, 7 Jun. 2016, Link.
  4. [4] Maximillian, J., Brusseau, M.L., Glenn, E.P., Matthias, A.D. “Environmental and Pollution Science (Third Edition), Chapter 25, Pollution and Environmental Perturbations in the Global System”. Science Direct, 1 Mar. 2019, Link.
  5. [5] “Global Greenhouse Gas Overview”. United States Environmental Protection Agency, 11 Apr. 2024, Link.
  6. [6] “U.S. Precision Agriculture Study Unveiled by AEM, AG Organizations”. Association of Equipment Manufacturers, 1 Feb. 2021, Link.
  7. [7] Sichugova, L., Fazilova D. “Soil Moisture Estimation Using Landsat˗8 Satellite Data: A Case Study the Karshi Steppe, Uzbekistan”. International Journal of Geoinformatics 18 (1):63-69, 7 Feb. 2022, Link.
  8. [8] Zhou, X., Zhang, J., Chen, D., Huang, Y., Kong, W., Yuan, L., Ye, S. “Remote Sensing of Vegetation Phenology in the Subtropical Forest of South China Using Landsat Data”. Journal of Plant Ecology, 1 Aug. 2022, Link.
  9. [9] Harris, M. “Understanding NDVI: An Overview”. National Oceanic and Atmospheric Administration (NOAA), 1 Jan. 2017, Link.
  10. [10] Zhang, Y., Liu, X., Zhang, L., Yang, X. “Applying the EVI Index to Remote Sensing Data for Crop Yield Prediction”. International Journal of Remote Sensing, 4 Sep. 2021, Link.

  11. Images & Web Design
  12. “Green Rice Farm.” PxHere, Accessed 19 Jul. 2024, Link.
  13. Manna, M. “Soil Moisture.”Artwork generated using Canva, 15 Jul. 2024, Link.
  14. Manna, M. “Chlorophyll Content.”Artwork generated using Canva, 15 Jul. 2024, Link.
  15. Manna, M. “Surface Temperature.”Artwork generated using Canva, 15 Jul. 2024, Link.
  16. Meyers, D. "Pile of Leafed Plants." Upsplash, 19 Jul. 2019, Link.
  17. “Digital Presentation Cartoon Illustrations.” Storyset, Accessed 22 Jul. 2024, Link.
  18. Eley, P. “Planet Earth.” Artwork generated using ChatGPT 4o, 4 Aug. 2024, Link.
  19. Chowdhury, M., Hammad, S. “Unique UI Components, Snippets, and Blocks for Bootstrap.” Ayro UI, 16 Apr. 2023, Link.
  20. “OpenAI.” ChatGPT, Jun. 2024 Version, Large Language Model, Link.

Meet Our Team

Pascual Eley

Pascual Eley

Data Scientist
Josh Fram

Josh Fram

Data Scientist
Maria Manna

Maria Manna

Data Scientist
Diego McDonald

Diego McDonald

Data Scientist
Cameron Yenche

Cameron Yenche

Data Scientist
Advisors
Advisors: Korin Reid & Ramesh Sarukkai