Empowering Rural Women Through AI for Sustainable Agriculture

Demonstrating the use of TEROS12 Soil Moisture sensor and entering data into logbook and KoboCollect app | Photo credit: Manjunatha G

Considering the frequent dry spells and erratic rains caused by intensifying climate change, adapting to monsoon variability is a foundational step toward resilience. It requires a fundamental shift in strategy, from maximising yield by expanding large-scale irrigation, to focusing on mitigating risk through protective irrigation; that is, by providing timely access to water needed to prevent crop failure during dry spells.

Effective soil moisture management is at the core of this protective strategy. Soil moisture, the water in the tiny spaces between soil particles, is a key source of water for crops, serving as a lifeline by enabling nutrient absorption and growth. When managed well, soil moisture can buffer crop yields from climate risks, such as long dry spells and weak monsoons, and act as a critical resource for smallholder farmers.

For effective management, measuring and monitoring soil moisture is crucial. The measurements can enable farmers and researchers to make decisions and understand climate trends. By tracking real-time moisture levels, farmers can transition from schedule-based irrigation to ‘just-in-time’ delivery, ensuring crops receive water precisely when they hit a stress point, conserving limited groundwater reserves.

 To accurately measure soil moisture variation within a given plot, multiple quantifications are required to capture the heterogeneity. However, conventional monitoring methods require expensive sensors for in-situ measurements and spot surveys, making it expensive and limiting scalability. Without obtaining data at scale, effectively managing large-scale irrigation and moisture deficits is challenging.

This obstacle can be overcome through satellite-based remote sensing soil moisture estimates, which offer a broader spatial scale. Nevertheless, the resolutions are very coarse, at 1 km or larger, making them difficult to use for individual smallholder farms and fields.

The role of these data collectors, many of whom are local women acting as citizen scientists, is fundamental to the success of this initiative. By providing high-quality, local sensor readings, they enable the ground-truthing necessary to validate and refine satellite-based AI models. These collectors bridge the gap between remote sensing and on-ground reality by providing thousands of data points across diverse crops and soil types. Their work directly improves the accuracy of national-scale soil moisture maps, which are essential for creating large-scale crop security plans. Ultimately, this effort supports the resilience and livelihoods of millions of rainfed farmers across India.

Leading Data Collection: Building Women’s Confidence and Social Status

Rural women have historically been excluded from technology-led initiatives due to systemic barriers and a lack of exposure. However, the AI for Soil Moisture Management project is placing women at its core. The women are trained to use the TEROS12 sensor and the KoboCollect data logging app, because of which they understand the importance of data quality and ethical technology handling.

By leading the data collection efforts, the women of Domalapalli gain far more than just technical proficiency; they undergo a deep personal and professional transformation. One of the most immediate changes is a surge in self-efficacy. As they master complex tools, these women gain confidence in being able to solve practical, real-world problems using technology. This expertise often places them in leadership roles within their community, where they frequently help their male peers navigate mobile applications or perform intricate sensor calibrations.

Beyond personal growth, the initiative also provides a vital economic opportunity. The women receive direct economic incentives tied to the reliability and scientific quality of the collected soil moisture data. This financial earning, paired with their new digital skills, significantly builds their social capital. It effectively dismantles long-standing stereotypes regarding technological roles of women in rural India, and sparks a sense of curiosity and possibility among the next generation of girls.

The women receive direct economic incentives tied to the reliability and scientific quality of the collected soil moisture data

Photo Credit: Manjunatha G

Ensuring Quality Via Strict Data Collection Rules

A typical day for a data collector begins at the time of the Sentinel-1 satellite overpass, at around 6 am. Work is dispatched via WhatsApp, evidence of rural digitisation’s rising importance. The collectors prepare their sensors, ensuring they are charged and ready for accurate readings. The quality of an AI model’s output is only as good as the quality of the input data. Therefore, to ensure the collected data is robust and accurately reflects the complex dynamics of the field, data collectors follow a detailed, scientific protocol.

At each field, they execute a precise ‘9-point’ sampling algorithm. This ensures they capture the variability of soil moisture by measuring at the roots and between rows. Armed with their notebooks and smartphones, they meticulously follow these steps:

• Insert the sensor and wait for stable readings before recording each data point.
• Alternate positions to capture the complexity of the soil-crop dynamics.
• Log every detail with field IDs and timestamps.
• Upload data using the KoboCollect app.

This enables the collection of high-quality input that can help transform local data to create national-scale maps. Through this, these women serve as data stewards who co-design the technology for real-world needs. They contribute to the protocol in the following ways:

• Identify contextual needs: Collectors identify where rigid scientific grids might fail due to local obstructions or unique field geometries.
• Suggest protocol tweaks: They suggest adjustments to the sampling process based on their on-ground experience, ensuring the technology is not merely imposed but is functional for smallholder farms.
• Perform operational troubleshooting: By mastering sensor calibrations, they help peers solve technical issues that could otherwise result in data gaps.

Accordingly, the technology becomes a shared tool for resilience, grounded in the meticulous work and local insights of these rural champions. This project creates a replicable model for gender inclusion that can be scaled to other villages. As more women step forward, a new pipeline is created for training future agritech leaders, field researchers, and data stewards across rural India. This is essential for an inclusive agricultural future, ensuring that the benefits of AI-driven innovation are shared across genders.

Cultivating a Sustainable and Inclusive Future

WELL Labs, Stanford, WASSAN, and GRAVIS aim to build the first public, open benchmark dataset of smallholder farm soil moisture in India. Scaling this effort to future-proof farms and reinforce community resilience requires expanding collector cohorts. Investing in digital literacy and providing ongoing support so all rural people can participate is, therefore, essential.
By centreing women as ambassadors of community-led science, crucially by recruiting and mentoring more women, technology and tradition meet to cultivate leadership, hope, and a new vision for equitable rural prosperity. This collaborative data collection does not just result in maps; it creates tangible use cases that transform how stakeholders manage water:

• For individual farmers: Accurate soil moisture predictions give farmers precise information on applying water at the right time to prevent crop failure.
• For local policy makers: Rapid drought identification enables government agencies to target investments in water-saving interventions more effectively.
• For researchers: Open benchmark dataset provides a foundation for the development of new agricultural tools tailored to the unique conditions of Indian smallholder farms.

The shift towards a risk-mitigation strategy based on protective irrigation, supported by AI and grounded in local women-led data collection, is a critical step towards building an inclusive and sustainable future for Indian agriculture.

Related Posts