SupPlant provides high-resolution forecasts, real-time insights, and irrigation automation to improve plant health for robust harvests with optimum water usage.
We recently spoke with SupPlant chief technology officer Revital Kremer about the company’s beginnings, its future, and why it decided to work with DataStax. Brimming with enthusiasm, she describes the company’s mission to help every farmer grow healthier plants through AI-powered irrigation, backed by real-time data using DataStax Astra DB.
1. Tell us about SupPlant and what sets it apart.
SupPlant is a world leader in precision agriculture. Farmers are under pressure to do more with less, particularly when it comes to water. We help farmers make informed decisions as they manage their crops by providing an additional real-time data source—the plants themselves.
Farmers rely on learned and earned experiences to manage their crops. But there are endless variables that make precision farming difficult. We gather data from each farmer’s crops and send it to the cloud, where artificial intelligence turns the data into real-time insights that support proactive farming, irrigation planner, and crop success.
There are 540 million farmers worldwide, of which there are 480 million small farmers. We want to help them all. Even the smallest farmer without sensors in the field can benefit from access to insights from SupPlant.
2. Let’s focus briefly on your journey and how it brought you to your position of impact at SupPlant. What personally motivated you to take on these challenges on such a broad scale?
Here is a cheat sheet to know if what you are doing is meaningful: do you like telling it to your/ other kids? If the answer is "yes", you are in the right place.
3. How does your technology work?
SupPlant’s system gathers data from sensors placed on the plant trunk, on the fruit, and in the soil, collecting and transmitting measurements in real time. SupPlant has accumulated data covering thousands of growing seasons across every weather condition. We process this data in real time to drive recommendations and irrigation commands.
4. Why is real-time data important to your customers?
Real-time data is critical for farmers, who rely on it for decision support to make better irrigation decisions every day. We identify trends by day and night, analyze patterns, and detect plant stress to support a proactive approach by farmers. We examine data in very short cycles, otherwise it will be too late for some farmers. And with our irrigation planner systems, adjustments are automatically made as needed across the farm and around the clock, which can be a huge advantage for farmers.
5. How do you use Cassandra?
Our data architecture has four key elements: raw and processed data in a managed Cassandra environment using DataStax Astra DB, metadata and system entities in MySQL, static collections and libraries in Elastic, and app cache in Redis.
Our Astra DB environment includes 1.5 billion data points collected from sensors. This represents 2,000 crop seasons from 2016 forward and includes 25 commercial crop types. Another 100 crop types are in development. The data is managed in a multi-tier architecture.
6. Why use a multi-tier architecture?
Data gathered from IoT devices must be tagged and cleaned for data integrity before it is used in models. We rely on a multi-tier approach using keyspaces to ensure we generate the best recommendations and instructions for each farmer. Users have access to all data tiers for time series, data assembly, or other purposes.
The first tier is the raw data keyspace, where data from IoT sensors, satellites, and APIs is collected every 30 to 60 minutes. The data is normalized and stored in a source-agnostic manner. Data includes trunk measurements, fruit size, soil moisture, and weather conditions.
The second tier is the daily summary keyspace, where data is processed into accumulations, statistics, and our own unique insights created daily. The data is processed in near real time and stored again.
The third and final tier is the features keyspace that handles AI model results, features, and predictions. Features and model results run in offline mode, producing additional insights.
7. Why did you start working with DataStax?
We chose DataStax Astra DB as the primary data store for our time-series IoT data because of its support for simple and cost-effective business continuity. A managed solution for Cassandra, Astra DB allows us to enjoy all the benefits of Cassandra while sparing us from managing the system. DataStax has the best support team ever, providing an excellent Database-as-a-Service (DBaaS) offering and helping us when we have architectural issues, encounter production challenges, or need general support.
With the help of DataStax, we’re enabling farmers to better listen to their plants, and manage their precious water resources in a climate-adaptive manner. This proactive and data-driven approach helps farmers prevent plant stress, reduce fruit loss, improve production, and maximize crop potential.
8. What advice would you give to any other enterprises or developers that are trying to win with data?
Invest time in data quality from day one. It will take double the time (or more) to do backward fix and fill-in the missing parts. Treat data with caution, especially when it comes to unreliable sources. Once you have enough data - let it tell you the story. Adopt a “data river” approach and build a data lake with constant data processing in several levels. This will help you utilize data more efficiently. Last but not least, your organization mindset must be set to data. Everyone must understand its importance and the value that can be gained out of it. This way people will continue to collect it, preserve its accuracy and use it as part of everyday work tasks.
Digital Champions are the builders of data-driven, high-growth businesses. They are the visionaries and driving forces in using real-time data and the cloud to deliver unprecedented value to their organizations. Hear how these Digital Champions like Revital Kramer are building powerful real-time applications that are making the difference.
