Virtual MOS Sensor Array Design for Ammonia Monitoring in Pig Barns.

Area of Science:

• Environmental Science
• Agricultural Engineering
• Sensor Technology

Background:

• Poor air quality in animal barns, particularly ammonia, poses significant respiratory health risks.
• Existing ammonia monitoring systems are often expensive and labor-intensive.
• Metal Oxide Semiconductor (MOS) sensors offer a cost-effective and scalable solution for air quality monitoring.

Purpose of the Study:

• To develop an improved ammonia detection method using MOS sensors and machine learning.
• To create a virtual sensor array for high-resolution spatial and temporal ammonia monitoring.
• To assess the effectiveness and economy of MOS sensors for continuous ammonia monitoring in pig barns.

Main Methods:

• Designed a virtual sensor array using a cyclic data-driven approach, integrating machine learning with solid-state sensor modeling.
• Utilized temperature-cycled operation of MOS sensors to enhance selectivity and create virtual sensor arrays.
• Implemented a filter membrane to improve accuracy and prevent sensor contamination.

Main Results:

• Demonstrated effective ammonia sensing in pig barns with a sampling rate of approximately 2/min and a range of 1-30 ppm.
• The developed method showed robustness, with a 10% increase in normalized RMSE when testing an unseen sensor module.
• A filter membrane successfully boosted accuracy and prevented data loss from contamination like flyspecks.

Conclusions:

• The BME688 MOS sensor is effective and economical for widespread, continuous ammonia monitoring in animal barns.
• The developed system enables accurate localization of ammonia sources, contributing to improved animal welfare.
• This approach offers a practical solution for real-time air quality management in agricultural settings.