A woman holding a pile of tomatoes.
CASE STUDY
How we helped 10 desert growers combat salinization.
Read more

Farms we've helped

Bee Sweet Citrus • Temalpakh Farms • Salton Sea Farms • Grimmway Cal Organic Farms • Bautista Organic Dates • Regulus Ranch • Sam Cobb Date Farms • Long Life Farms • Wong Farms • Aziz Farms •
Bee Sweet Citrus • Temalpakh Farms • Salton Sea Farms • Grimmway Cal Organic Farms • Bautista Organic Dates • Regulus Ranch • Sam Cobb Date Farms • Long Life Farms • Wong Farms • Aziz Farms •
Salton Sea
LOCATION
Desert orchards
TYPE
Specialty crops
FOCUS
Overview
The problem

For many orchards in southeastern California, irrigation runoff from Coachella and Imperial Valleys has resulted in a salinity that's currently 50% greater than that of the ocean. This high salt concentration disrupts the soil structure, reducing water infiltration and leading to waterlogging, root zone toxicity, and interference with nutrient uptake by plants.

The goal

Drip irrigation, while effective in conserving water, inadvertently exacerbates salinity issues by causing salts to accumulate in the root zone. This led to an average 14% decrease in soil fertility, reducing the nutrient availability for crops and ultimately impacting tomato yields by up to 10% due to salinity and disease stresses.

For specialty crop farmers, this adversely affects water balance regulation within plants, causing leaf burn, reduced fruit quality, and yield losses of up to 30%. Organic farmers face additional challenges, as the salinity limits the effectiveness of natural soil amendments like compost and green manure, complicating efforts to maintain soil fertility and health without synthetic fertilizers.

The primary business objective of Agsight was to increase the yield of 10 tomato farms by 10% over the next 2 growing seasons. This would be achieved by monitoring and adjusting irrigation schedules through near real time machine learning algorithms that ensured optimal water use to reduce soil salinity levels by at least 20% and water waste by 15%.

01
Auditing the current experience

We engaged deeply with farmers in the San Bernardino, Riverside, San Diego, and Imperial counties to audit their current experience with irrigation and salinity management. Through this process, several key usability issues emerged. Farmers consistently highlighted the complexity of managing drip irrigation systems due to inconsistent soil moisture readings, inaccuracies in monitoring tools, and the time-consuming nature of manually adjusting irrigation schedules. This often led to under or over-irrigation, reducing yields by ≈11-14% annually.

Interviews, field studies, and research.
6 field studies conducted, 8 competitors audited, and 10 farmer interviews.
02
Asking the right people the right questions

When conducting field studies, we conducted in-depth interviews among 10 farmers and agronomists to understand their pain points related to using existing technologies for irrigation scheduling, water management, sensor data interpretation, and salinity issues. Based on data from existing sensors across a representative subset of 6 farms, we were able to use data on soil moisture, electrical conductivity (EC) for salinity assessment, and climatic metrics to gain insights into variations in soil conditions and irrigation effectiveness over time.

How do you currently manage irrigation?
How do you deal with high soil salinity levels?
How do you receive weather forecasts and integrate them into irrigation planning?
What features do you find useful in your current irrigation tools?
How do you interpret data from sensors?
How easy is it for you to understand and act on salinity data?
How do you currently manage irrigation?
How do you deal with high soil salinity levels?
How do you receive weather forecasts and integrate them into irrigation planning?
What features do you find useful in your current irrigation tools?
How do you interpret data from sensors?
How easy is it for you to understand and act on salinity data?
What kind of personalized recommendations would you find most helpful?
How do you currently monitor salinity levels?
What information do you need to better manage soil salinity?
What methods have you tried to mitigate salinity in your soil?
How do you determine when to apply soil amendments to combat salinity?
How do you test your soil?
What kind of personalized recommendations would you find most helpful?
How do you currently monitor salinity levels?
What information do you need to better manage soil salinity?
What methods have you tried to mitigate salinity in your soil?
How do you determine when to apply soil amendments to combat salinity?
How do you test your soil?
03

Insights and opportunities

Translate esoteric data into actionable steps.

Farmers find it challenging to understand and act on salinity data to optimize their irrigation practices. Real-time data from sensors often fails to translate into clear, actionable steps to mitigate the impacts of salinity.

Prioritize personalized data.

There's a significant demand for personalized recommendations that integrate salinity data with local soil and weather conditions to guide precise irrigation and soil management practices.

Make agtech systems more accessible.

The complexity of existing platforms and the need for installing and maintaining fragile sensors overwhelms farmers, who prefer simpler, more intuitive interfaces that focus on essential data and provide clear, actionable recommendations.

A farmer and his wife posing in front of their tractor in front of their fields.
04
Our solution
Soil algorithms

The complexity of existing platforms and the need for installing and maintaining fragile sensors overwhelms farmers, who prefer simpler, more intuitive interfaces that focus on essential data and provide clear, actionable recommendations.

An irrigation canal.
Where we landed

Read the technicals

We created the integrated salinity management system (ISMS), which combines all 3 approaches in a way that optimizes both irrigation and soil fertility to sustainably increase yield.
Using machine learning models trainedn data webscraped from USDA NCRS, NASA, NOAA, and USGS, we used variable rate irrigation (VRI) eachamount of water based on its unique needs.
The model suggests soil amendments to areas it predicts has high salinity levels, crop rotation schedules that include salt-tolerant crops to break salinity cycles, and optimal times for leaching.
A farmer operating his tractor.
A farmer standing in his orchard with a grocery bag filled with produce.
We gathered data from USDA NRCS, NASA, NOAA, and USGS. This included soil composition data, historical salinity levels, weather patterns, and crop-specific iUsing aated scripts, we collected and processed large datasets from these sources to train our M models.
A farmer plucking an apple from an orchard tree.
Using the ML models trained on historical and real-time data, the ISMS implements VRI to optimize water distribution to calculate the precise amount of water needed for each section of the field based on irrigation, nearby saltwater sources, salinity, forecasts, and crop growth stage.
Because we use ML to estimate agricultural data, ISMS does all of this without the need of ground-truth data from sensors with an accuracy of 98.6%. This provides the benefits of personalization without the infrastructural, intellectual, and financial barriers associated with installing sensors.
Validation
To validate the ISMS, we conducted extensive field trials across multiple designed to assess the system's ability to mitigate salinity stressoptimize irrigation practices
Prior to the trials, farms were equipped with sensors and traditional irrigation practices (control group) and ISMS (experimental group). Results showed that ISMS optimized VRI, recommended soil amendments, and planned crop rotations much more effectively.
9.5
%
average yield increase
12
%
decrease in water use per acre
17
%
decrease in soil electrical conductivity
16
%
increase in marketable (No. 1) crops
“I never thought in my wildest dreams this type of difference could exist in the microclimate of one block with two different approaches.”
A farmer with gray hair holding a basket with peaches in an orchard.
Matthew Watkins, Bee Sweet Citrus
A phone holding a phone screen displaying salinity levels in a field.
33
%
fewer instances of physiological disorders (e.g., Blossom-end rot)
20
%
lower nitrate concentration in leachate
A dirt trail in a farm. On its left is a field of withering crops and on its right is a dryland orchard.
"Agsight has been amazing. It’s super easy to use, it’s intuitive, and pulls in more data than we even thought was possible, from soil health and irrigation strategies. It's quantifying everything that’s happening with our soil over time."
A farmer pruning a dead tree branch.
John Frank, Temalpakh Farm
Real-time maps

We integrated a novel real-time salinity mapping feature into our platform, which aggregated webscraped data to create detailed salinity maps with localized information about salinity levels in different parts of their fields.

A hexagon resting on a corn on the cob.
Salinity forecasts

By analyzing patterns in soil moisture, salinity levels, and weather conditions, our system predicted periods of high salinity stress to suggest preventative actions, such as adjusting irrigation or applying soil amendments.

A farmer's foot hoisting itself onto his tractor.
ISMS

Experimental group

Several rows of prospering trees in an orchard.
Traditional

Control group

Several rows of withering trees in an orchard.
Two miniature cartoon figures with party poppers shooting produce.

RELEASE

Broader release scheduled for 2025

This feature is currently in beta testing and undergoing a larger scope of validation to ensure it's perfect when released!

NEED A HAND?

Need help transforming your farm?
We'd love to talk!
Let's talk!