Live Demo - INAPINE Smart Agricultural System

1 System Overview

The INAPINE Smart Agricultural System integrates hardware and software components into a unified monitoring solution. Solar-powered IoT traps are deployed across fields, capturing images and environmental data continuously — day and night — without requiring grid electricity or manual intervention.

Solar Powered

Autonomous energy harvesting with battery backup for 72h offline operation

AI Detection

Computer vision model identifies 50+ pest species with 94.3% accuracy

Cloud Dashboard

Real-time data visualization accessible from any device, anywhere

2 Device Workflow

Each INAPINE smart trap follows an automated end-to-end workflow, requiring zero manual intervention once deployed in the field.

Step 1 — Image Capture

High-resolution camera triggers every 30 minutes (configurable). IR night-vision ensures 24/7 operation. Images stored locally in case of connectivity loss.

Step 2 — On-Device AI Inference

Edge AI model (TensorFlow Lite) runs pest detection locally on the Raspberry Pi CM4. Results include species ID, count, confidence score, and bounding boxes.

Step 3 — Data Transmission

Compressed data packets sent via 4G/LTE MQTT over TLS to the cloud broker. Sensor readings (temperature, humidity, soil moisture) included in each payload.

Step 4 — Dashboard & Alerts

Cloud backend processes data, generates trend charts, and triggers SMS/email alerts when pest thresholds are exceeded. All data visible in real-time on the dashboard.

3 Live Metrics Dashboard

The following metrics reflect the current state of the active INAPINE prototype deployment at Institut Agro Montpellier.

INAPINE Field Monitor — Live Connected

Pests Detected

847

↑ +12 today

Temperature

24°C

Optimal range

Soil Moisture

67%

↓ Needs water

Battery Level

92%

Solar charging

Species Breakdown (Last 7 Days)

Medfly (Ceratitis capitata)42%

Aphids (Aphis gossypii)28%

Whitefly (Bemisia tabaci)18%

Other species12%

Active Traps Status

Trap A-01 (North Field) Online

Trap A-02 (South Field) Online

Trap B-01 (Greenhouse 1) Syncing

Trap B-02 (Greenhouse 2) Online

4 Solar Energy System

Each INAPINE trap is powered entirely by a 20W monocrystalline solar panel with a 10,000mAh LiFePO4 battery. This architecture eliminates the need for grid power, allowing deployment in the most remote agricultural areas.

Energy Budget (Daily)

Solar generation (avg) +80 Wh

Camera + AI inference -28 Wh

4G modem (transmission) -18 Wh

Sensors + MCU -8 Wh

Net daily surplus +26 Wh

Offline Resilience

72-hour battery autonomy

Operates fully during cloudy periods or at night

Local data buffering

Up to 30 days of readings stored on SD card

Auto-sync on reconnect

Buffered data pushed automatically when connection resumes

5 Smart Irrigation Control

INAPINE integrates capacitive soil moisture sensors to feed AI-driven irrigation recommendations. The system correlates pest activity with moisture data — excessive humidity is a known attractor for several pest species.

67%

Current Moisture

Field sensor avg — below optimal 70–80% range

18L/m²

Recommended Dose

AI-computed irrigation volume for next cycle

06:30

Next Scheduled

Tomorrow morning — optimal evaporation window

6 Prototype Deployment — Institut Agro Montpellier

The first INAPINE prototype has been deployed at Institut Agro Montpellier (Domaine du Chapitre, 34750 Villeneuve-lès-Maguelone) since January 2025. The pilot covers 4 traps across two greenhouse zones and two open-field areas.

4

Active Traps

2 ha

Coverage Area

14mo

In Operation

94.3%

Detection Accuracy

Research Validation

Results from the pilot are being documented for publication in the Journal of Agricultural Technology and Innovation. The dataset covers 14 months of continuous pest monitoring across Mediterranean crop varieties.