Introduction: In the context of occupational biosecurity, the daily processing of biological samples such as feces and urine in microbiology laboratories poses an environmental risk hypothesis due to the potential emission of gases such as CO2 derived from organic decomposition and urea hydrolysis, the accumulation of which affects staff well-being and performance; therefore, given that previous literature lacks specific IoT systems for this environment integrating NDIR sensors (K30) and environmental compensation via BMP280 and SGP40, this project proposes the design and development of a continuous monitoring device to transform this invisible risk into objective data that enhances the surveillance of air quality and working conditions in the laboratory.
Methods: To measure CO2 concentrations in a microbiological laboratory, the SenseAir K30 NDIR sensor was used; furthermore, a calibration algorithm based on environmental variables measured with the BMP280 sensor, utilizing a quadratic polynomial regression, was incorporated. The processing unit is a Lilygo TTGO SIM7000G board, based on an ESP32 microcontroller with an integrated cellular chip.
Results and Discussion: To validate the performance of the environmental monitor, comparative tests were conducted between the measurements of the K30 sensor and those of the reference equipment, CO₂ Monitor-1200P SA12002104191975, in a laboratory over 1 week, recording a contamination peak of 2,887 ppm at the highest workload. Without calibration, the device showed differences of 50–100 ppm compared to the reference, and after integrating the calibration algorithm, these were reduced to 10–20 ppm. This improvement is consistent with the results of Parkavi (2026), in which multivariate correction for pressure, temperature, and water vapor reduced the RMSE from values of 10–21 ppm to 1.7–4.3 ppm with four weeks of training data.
Conclusions: A correlation was observed between increased urine volume and CO2 concentration; better results were obtained in closed environments than in open ones due to lower environmental variability; finally, a longer calibration period is suggested to reduce the RMSE.