Skip to main content

Agriculture / AgriTech

Smart agriculture: Solar-powered LoRaWAN sensor network for precision irrigation

Design of a solar-powered precision-agriculture IoT system — soil and weather sensor nodes, LoRaWAN gateways, irrigation models and a multilingual farmer app — piloted with 100 nodes on 200 acres, where the pilot measured 40% higher soybean yield and 35% lower water use, before a cooperative-wide rollout.

What we did: We designed the solar sensor-node hardware and firmware, the gateway and cloud setup, the irrigation and fertiliser models, and the Marathi/Hindi/English Android app.

Agricultural cooperative, Maharashtra, India10 months

Published Last reviewed:

40%
Increase in crop yield
Measured in pilot
35%
Reduction in water usage
Measured in pilot
200
Acres in pilot
100
Sensor nodes in pilot
Illustrative photo: a solar-powered soil sensor in a crop field with drip irrigationIllustrative image

The challenge

An agricultural cooperative in Maharashtra managing 5,000 acres across 200+ farms needed a precision agriculture solution to optimize water usage, improve crop yields, and provide actionable insights to farmers with varying levels of technical literacy.

Water scarcity

The region faced severe groundwater stress, and traditional flood irrigation wasted a large share of the water applied.

Impact: ₹3Cr annual water costs (~$350K)Client-reported baseline

Inconsistent yields

Crop yields varied by 50% across similar plots due to lack of data-driven decision making. Farmers relied on intuition and traditional practices.

Impact: 30-40% below potential yieldClient-reported baseline

No connectivity

Most farms had no cellular coverage or reliable power supply. Any IoT solution needed to work completely off-grid with long-range communication.

Impact: Zero existing infrastructureClient-reported baseline

Diverse user base

Farmers ranged from tech-savvy young graduates to elderly farmers with no smartphone experience. The system needed to serve all user types.

Impact: Multilingual, voice-based UI requiredClient-reported baseline

Impact figures describe the situation before the project, as the client reported it to us; we did not measure them.

Our solution

We designed a precision agriculture system including solar-powered sensor nodes, LoRaWAN networking, data-driven irrigation and fertiliser recommendations, and a multilingual mobile app with voice interface for farmers.

System architecture

Fully off-grid architecture designed for remote agricultural environments with no existing infrastructure.

Field sensor layer

  • Soil moisture sensors at multiple depths
  • Soil NPK nutrient sensors
  • Weather stations (temperature, humidity, rainfall, wind)
  • Leaf wetness and canopy sensors
  • Water flow meters for irrigation monitoring

Communication layer

  • LoRaWAN network, IN865 band (up to 15 km line of sight)
  • Solar-powered gateway nodes
  • Satellite backhaul for remote areas
  • Local data buffering for connectivity gaps

Intelligence layer

  • Crop-specific ML models for irrigation scheduling
  • Weather forecasting integration
  • Pest and disease risk prediction
  • Yield estimation models
  • Fertilizer recommendation engine

Farmer interface

  • Android app in Marathi, Hindi, English
  • Voice-based interaction for farmers who prefer not to read
  • SMS alerts for critical events
  • WhatsApp integration for recommendations
  • Community dashboard for cooperative

Custom hardware design

MCU
STM32L072 (ultra-low power)
Radio
Semtech SX1262 (LoRaWAN, IN865: 865–867 MHz)
Solar Panel
2W polycrystalline
Battery
3.2 V 6000 mAh LiFePO4
Sensors
Capacitive soil moisture, I2C NPK
Enclosure
IP66 UV-resistant ABS
Operating Life
5+ years maintenance-free (design target)

Firmware features

  • Adaptive sampling based on crop growth stage
  • LoRaWAN Class A networking
  • Local data storage for 30 days offline operation
  • Solar power management with MPPT
  • Self-diagnostics and remote health monitoring
  • OTA firmware updates over LoRa

AI-powered recommendations

We developed crop-specific ML models trained on local agricultural data to provide actionable recommendations. Irrigation and yield-estimation accuracy were measured on the pilot crop cycle; the pest-outbreak figure is a design target, because one season on 200 acres has too few outbreaks to measure it.

Irrigation scheduling

Ensemble (Random Forest + LSTM)

92% water need prediction

Measured in pilot

Pest risk prediction

Gradient Boosting classifier

87% pest outbreak prediction

Design target

Yield estimation

CNN on satellite + field data

±8% yield prediction

Measured in pilot

Fertilizer recommendation

Rule-based + ML hybrid

Crop-specific NPK optimization

Implementation timeline

Phase 1: Research & design

8 weeks
  • Field visits and farmer interviews
  • Soil type and crop analysis
  • Network coverage survey
  • Hardware and software architecture

Phase 2: Pilot development

10 weeks
  • 100-node pilot deployment (200 acres)
  • Gateway and cloud infrastructure setup
  • Mobile app development
  • Initial ML model training

Phase 3: Pilot validation

16 weeks
  • One full crop cycle monitoring
  • Model refinement with local data
  • Farmer feedback incorporation
  • Reliability and durability testing

Phase 4: Scale deployment

12 weeks
  • Sensor node production for the cooperative-wide rollout
  • Farmer training programs
  • Cooperative dashboard launch

Results & impact

These results were measured on the 200-acre pilot (100 nodes) over one crop cycle. Financial, income and return-on-investment estimates are not published.

Video poster: “40% higher soybean yield” and “35% less water”, labelled Measured in pilot, over an illustrative photo of a solar-powered sensor pole in a drip-irrigated soybean field at dawn.

Solar LoRaWAN sensors for precision irrigation25 s · no sound

Video text
  1. 40% higher soybean yield. 35% less water. (Measured in pilot.)
  2. Case study · Agricultural cooperative, Maharashtra. No cellular coverage. No reliable power. Across most farms of a 5,000-acre cooperative.
  3. An off-grid sensor node. IP66 enclosure. 30-day local storage. Power: Solar panel 2 W → MPPT charger → LiFePO₄ 3.2 V · 6000 mAh → Regulator. Data: NPK sensor (I²C) and capacitive soil moisture → STM32L072, LoRaWAN Class A → SPI → SX1262, LoRa · IN865.
  4. Pilot: 100 nodes · 200 acres · one crop cycle. Farmer app in Marathi, Hindi and English.
  5. 40% higher soybean yield. 35% less water. 25% lower fertiliser cost per acre. (Measured in pilot.) Pilot plots, one crop cycle.
  6. Solar LoRaWAN sensors for precision irrigation. Rapid Circuitry. Discuss your product → rapidcircuitry.com/contact. Reply within one business day · Mutual NDA on request.
  7. The photo is an illustrative image.

How to read the labels: Measured in pilot = measured during the project's pilot deployment; Design target = an engineering target, not a field measurement; Client-reported baseline = the client's own figure for the situation before the project; we did not measure it. Figures without a label are specifications or project scope.

Crop yield increase

Measured in pilot

40%

Higher soybean yield on the pilot plots

Water consumption

Measured in pilot

35%

Lower water use on the pilot plots

Fertilizer usage

Measured in pilot
Before:₹8,500/acre (~$100/acre)
After:₹6,375/acre (~$75/acre)
25% reduction

Technologies used

  • STM32L072
  • LoRaWAN
  • SX1262
  • Solar MPPT
  • TensorFlow
  • React Native
  • AWS IoT
  • PostgreSQL
  • TimescaleDB
  • Python

Related case studies

Manufacturing / Heavy Industry

Industrial IoT: Predictive maintenance sensors and edge AI for a steel plant

Infrastructure / Civil Engineering / Smart Cities

Battery-free IoT: Energy-harvesting sensor nodes for structural health monitoring

Related services, industries and guides