# Development of an Internet of Things–based real-time greenhouse gas and weather monitoring system for precision dairy farming

## TL;DR

A low-cost IoT sensor network was developed to monitor greenhouse gases and weather in real-time on dairy farms, helping reduce environmental impact.

## Contribution

A scalable, cost-effective IoT-based system for real-time monitoring of dairy farm greenhouse gas emissions and environmental conditions.

## Key findings

- Methane emissions were highest in dry cow pens compared to feed bunks and freestall beds.
- Ammonia emissions were highest in freestall beds, influenced by housing and manure practices.
- Sensing nodes showed moderate correlation with gold-standard respiration chambers for methane.

## Abstract

Summary: An Internet of Things-based wireless sensor network was developed and validated to monitor greenhouse gas emissions and environmental conditions on dairy farms. Cost-effective sensing nodes measuring methane, carbon dioxide, and ammonia were first calibrated using standardized gases and then validated against gold-standard open-circuit respiration chambers with lactating cows. The calibrated sensing nodes were subsequently deployed across different areas of a dairy farm, including feed bunks, freestall beds, and dry cow pens, to capture real-time variations in spatial and temporal greenhouse gas emissions. This scalable solution enables continuous real-time environmental monitoring. The resulting data may support precision dairy management practices aimed at assessing and mitigating the environmental footprint of dairy production systems.

Summary: An Internet of Things-based wireless sensor network was developed and validated to monitor greenhouse gas emissions and environmental conditions on dairy farms. Cost-effective sensing nodes measuring methane, carbon dioxide, and ammonia were first calibrated using standardized gases and then validated against gold-standard open-circuit respiration chambers with lactating cows. The calibrated sensing nodes were subsequently deployed across different areas of a dairy farm, including feed bunks, freestall beds, and dry cow pens, to capture real-time variations in spatial and temporal greenhouse gas emissions. This scalable solution enables continuous real-time environmental monitoring. The resulting data may support precision dairy management practices aimed at assessing and mitigating the environmental footprint of dairy production systems.

•Sensing nodes showed moderate correlation with a gold-standard respiration chamber.•The sensing network captured spatiotemporal greenhouse gas emission variations.•Emissions may be linked to cattle density, housing, and manure management practices.•Sensing nodes offer a cost-effective and scalable solution for emissions monitoring.

Sensing nodes showed moderate correlation with a gold-standard respiration chamber.

The sensing network captured spatiotemporal greenhouse gas emission variations.

Emissions may be linked to cattle density, housing, and manure management practices.

Sensing nodes offer a cost-effective and scalable solution for emissions monitoring.

The environmental footprint of dairy production is one of the most pressing challenges faced by the industry globally. Our study aimed to develop and validate a cost-effective sensing solution for real-time monitoring of dairy farms' GHG emissions and microclimatic conditions. Each of the integrated sensing nodes was equipped with carbon dioxide (CO2), methane (CH4), and ammonia (NH3) gas sensors, along with an all-in-one weather sensor. Sensing nodes were validated against gold-standard measurements using open-circuit respiration chambers with individual cows under controlled conditions. The CH4 emissions (133.0 ± 22.5 ppm, mean ± SD) showed an overall correlation (r = 0.46) with the gold-standard respiration chamber (166.0 ± 32.8 ppm) across all 3 d. However, the correlation changed over time, with a strong correlation on d 1 (r = 0.62), a moderate correlation on d 2 (r = 0.35), and a weak correlation on d 3 (r = 0.11). In contrast, sensor node quantified CO2 emissions (905 ± 779 ppm) showed a weaker correlation (r = 0.019, 2,461 ± 346 ppm), indicating the need for further improvements to the sensing node. A wireless network of calibrated sensing nodes was deployed in 3 different locations within a dairy farm: dry cow pen (DCP), feed bunk (FB), and freestall beds (FSB) at a research dairy farm. The CH4 emissions were greater in the DCP (12.5 ± 6.65 ppm) compared with FB (2.80 ± 0.61 ppm) and FSB (2.34 ± 0.62 ppm). The CO2 emissions at the FB were greater (1,498 ± 1,020 ppm) compared with the DCP (534 ± 222 ppm) and FSB (724 ± 517 ppm). The NH3 emissions were highest in the FSB (4.24 ± 0.91 ppm) compared with DCP (2.93 ± 1.35 ppm) and FB (1.10 ± 0.44 ppm). The differences in GHG emissions across the different areas of the dairy farm may be influenced by ambient temperature, humidity, housing conditions, and manure management practices. Our sensing nodes may provide a low-cost, scalable sensing network that can offer a practical solution for continuous GHG monitoring.