SIP5
IoT-assisted compost monitoring that accelerates maturity, improves quality, and reduces manual work
Sector
Arable/Cross-Sector / Tomatoes, Cucumber & Lettuce
Country
Cyprus
SIP Leader
EMBIO Diagnostics
Participants
7 farmers + 1 advisor
The challenge
Manual monitoring cannot keep up with compost variability
Open-pile composting is highly dependent on experience and frequent manual checking. Temperature, moisture, and pH are typically assessed by hand, requiring regular site visits, subjective interpretation, and reactive corrections that often come too late to prevent quality loss.
SIP5 set out to replace manual monitoring with continuous sensing and automated irrigation, giving compost producers a consistent, data-driven process that reduces labour, improves nutrient stability, and supports traceability from start to finish.
How the system works
Continuous sensing, automated irrigation, on-site validation
Wireless sensor nodes placed inside the compost pile continuously measure temperature, moisture, and pH. When moisture drops below a defined threshold, motorized irrigation valves activate automatically, adding water without any manual intervention. Weather forecasts are integrated to avoid watering before expected rainfall.
Compost quality is assessed using the B.E.L.D device, which performs NPK nutrient analysis directly at the pile. This validated measurement complements the sensor monitoring, giving producers a complete picture of both the process conditions and the final compost quality.
The system was deployed in two modes alongside a manually managed control pile: a Cloud mode that transmits sensor data via LoRaWAN and 4G for centralised visualisation; and an Edge mode that processes data locally with an offline B.E.L.D app, requiring no internet connectivity.
Wireless sensor nodes measure temperature, moisture, and pH inside the pile continuously.
System identifies deviations and triggers automated irrigation via eValves.
B.E.L.D device performs on-site NPK analysis to validate compost quality at key intervals.
Dashboard (Cloud) or local app (Edge) provides full process history, alerts, and traceability.
Cloud and Edge, equally recommended
Both modes received “definitely yes” for wider adoption in Year 1 – the only SIP where neither mode is preferred.
CLOUD MODEL
Sensor data transmitted via LoRaWAN and 4G to a cloud IoT platform. Supports remote monitoring, centralised data storage, and cloud-based B.E.L.D NPK analysis. Real-time alerts and visualisations accessible from any device.
Best for: producers who prioritise remote access, centralised data management, and cloud-based NPK analysis.
Data processed locally on an on-site mini-PC running open-source ThingsBoard. Offline B.E.L.D app for local NPK analysis without internet. More energy-efficient and fully open-source.
Best for: producers who prioritise data openness, open-source flexibility, or operate in low-connectivity environments.
What farmers said
After Year 1, participating farmers, compost producers, and advisors shared their experience with the IoT-assisted compost monitoring system in both Cloud and Edge modes.
Both modes: "Definitely recommended" for wider adoption
Both the Cloud and Edge modes received the strongest possible recommendation for wider adoption — "definitely yes" in both cases. This reflects the genuine practical value of IoT-based compost monitoring as an approach, regardless of which mode is chosen.
Training is essential for both modes to scale
Adoption willingness showed moderate resistance for both modes, an expected result given that IoT-assisted composting represents a significant change in practice. Digital skill requirements were rated as high for Cloud and high-to-moderate for Edge, highlighting the importance of structured training when scaling.
Edge wins on openness. Cloud wins on remote access.
The Cloud mode enables remote access to process data and is seen as a contributor to regional economic growth. The Edge mode is valued for its reliability, open-source architecture, and ability to operate effectively in low-connectivity environments. Data stays local and the software is fully customisable.
“Better compost, less guesswork. Both modes deliver.”
Year 1 headline results
months to maturity
Both Cloud and Edge accelerated compost maturity to 3 months. The manually managed control pile took longer and produced lower nutrient stability.
less manual labour
Cloud reduced manual monitoring and watering by approximately 30%. Edge achieved 28% reduction. Both eliminated the need for frequent manual moisture checking.
compost yield
Edge achieved the highest mass yield at 32.1%. Cloud delivered 30.7% with the highest nutrient stability. Control pile: 26.7%. Both significantly outperformed manual management.
Year 1 results: full breakdown
Cloud and Edge, side by side
Results are organised across 4 dimensions. Each compares the Cloud and Edge deployment mode side by side. Dimension 4 is represented above in the section “What farmers said.”
| Metric | Cloud | Edge | What this means |
|---|---|---|---|
| Composting speed and maturity | Cloud: 3-month duration. Most balanced maturity profile. Highest nutrient stability overall. | Edge: 3-month duration. Fastest decomposition rate. Slightly less stabilised than Cloud. | Both accelerate maturity vs manual control pile. Cloud prioritises quality; Edge prioritises speed and volume. |
| Reduction in manual labour | Cloud: approximately 30% reduction. Automated irrigation prevented extra manual watering. | Edge: approximately 28% reduction. Similar monitoring savings. | Both modes meaningfully reduce manual effort. The sensor-driven approach is the primary driver. |
| Overall production efficiency | Cloud: 158 kg final compost — 30.7% yield. Highest nutrient quality and stability. | Edge: 165 kg final compost — 32.1% yield. Highest mass output. | Edge maximises volume. Cloud maximises quality. Control pile: 26.7%. Both significantly outperform manual management. |
| Metric | Cloud | Edge | What this means |
|---|---|---|---|
| Network utilisation | Cloud: approximately 0.0128 GB per week per compost pile via LoRaWAN and 4G. | Edge: zero ongoing network transfer. Data collected and stored locally. | Edge eliminates ongoing network dependency. Cloud requires continuous connectivity and IoT platform subscription. |
| Computing resources | Same hardware footprint in the field. Edge handles hundreds of sensor nodes locally. Same computing requirements for both modes. Edge mini-PC (Asus NUC i7) handles multiple compost piles at 0-2% CPU steady-state. Cloud CPU metrics not available from pilot period. | ||
| Energy for data processing | Cloud B.E.L.D: ~0.8% battery per NPK analysis (41.9 mAh). | Edge B.E.L.D: ~1.0% battery per NPK analysis (50.7 mAh). | Both B.E.L.D configurations are energy-efficient for field use. Difference is minimal. |
| Total energy for farming operation | 14 visits x 5.8 kWh = 81.2 kWh transport energy. Same for both modes. 14 field visits required for both Cloud and Edge (vs 35 without ADS). Approximately 40% reduction in transport energy. Same number of visits in practice for both modes. | ||
| Metric | Cloud | Edge | What this means |
|---|---|---|---|
| Data formats | Same data collection and export for both modes. Both modes: time-series key-value format. OCSM ontology applied. CSV export available. Integrated with OpenAgri Farm Calendar via REST API. | ||
| Networking standards | Cloud: LoRaWAN for sensor collection. 4G for internet and data transmission. | Edge: LoRaWAN for sensor collection. 4G VPN for remote GUI access when needed. | LoRaWAN used in both modes. Edge uses VPN to allow remote interface access. |
| Software switchability | Cloud: datacake.co is proprietary. No source code access. Not switchable. | Edge: open-source ThingsBoard on local device. Fully customisable. Source code accessible. | This is the most significant D3 difference. Edge is fully open. Cloud is locked to the datacake.co platform. |
| Farm data switchability | Same data portability for both modes. Data downloadable from both platforms. Not locked to manufacturer format. CSV export available. | ||
| Hardware switchability | Same for both modes. Any LoRaWAN-compatible sensor. No vendor lock-in. Scalable to hundreds of piles in rural areas. | ||
OPENAGRI OS-BASED SERVICES USED
The following OpenAgri OS-based services are integrated in SIP5:
- Farm Calendar (centralised logging of sensor measurements, irrigation events, and sampling activities)
- Weather Data Service (context-aware irrigation decisions based on forecasted conditions)
- Reporting Service (generation of compost batch reports combining sensor data and management records)
PILOT DETAILS
- SIP Leader: EMBIO Diagnostics
- Partners: GSC GR, Farmers Fresh (3rd party)
- Sector: Arable / Cross-Sector (tomatoes, cucumber, lettuce)
- Country: Cyprus
- Target group: Farmers, compost producers, advisors