Testing an open-source multi brand sensor node to monitor variability of environmental conditions inside a greenhouse

Authors: Fletcher, Reginald; Fisher, Daniel

Submitted to: Agricultural Sciences
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 3/6/2021
Publication Date: 3/9/2021
Citation: Fletcher, R.S., Fisher, D.K. 2021. Testing an open-source multi brand sensor node to monitor variability of environmental conditions inside a greenhouse. Agricultural Sciences. 12(3):159-180. https://doi.org/10.4236/as.2021.123011.
DOI: https://doi.org/10.4236/as.2021.123011

Interpretive Summary: The agricultural community is interested in using miniature sensing technologies and open-source tools for precision agriculture, for monitoring greenhouse environmental conditions, and for checking soil moisture. Users are often interested in testing more than one sensor brand to determine how they compare and work with other sensor brands. To monitor environmental conditions in a greenhouse, scientists from the USDA-ARS Crop Productions Systems Research Unit in Stoneville, MS built a miniature integrated sensor node for simultaneous measurements of temperature, relative humidity, light intensity, and atmospheric conditions. The environmental monitoring system also allows the user to compare temperature and relative humidity readings of sensors designed by different manufacturers. Sensor measurements indicated that environmental conditions inside a greenhouse can vary considerably throughout the day and across physical locations within the greenhouse. Researchers or other users who assume or require uniform environmental conditions within a greenhouse can verify the uniformity, or measure the variability, of conditions using the relatively low-cost, open-source monitoring system described here.

Technical Abstract: A research project was undertaken to collect data to study the variability in environmental parameters inside a greenhouse. The specific objectives of the project were to 1) develop the network of open-source sensor nodes, 2) evaluate the performance of the individual sensors, and 3) quantify the spatial variability of environmental parameters within the greenhouse. The sensor system consisted of a sensor node equipped with three temperature and relative humidity sensors, one light-level sensor, one barometric pressure sensor, AA batteries, and a microcontroller board with a built-in radio to transfer the data wirelessly. The sensors were controlled with open-source technology. Twelve sensor nodes were fabricated and placed at different locations in a greenhouse to evaluate variability in sensor location and environmental parameters. Data collected during February 2019 were used to test the sensors. Heatmaps were employed to assess the variability of the measurements. Variability in greenhouse temperature, relative humidity, and light level conditions was identified with the sensor system. Overall, environmental measures based on time of day appeared to be a better grouping mechanism for analysis than sensor location in the greenhouse. Similar patterns were observed between the different sensor manufacturer’s heatmaps for the temperature sensors and relative humidity sensors. This study provided a protocol for developing the inexpensive multi-sensor sensor node and showed that automated measurements obtained with the system could help monitor variation in a greenhouse setting. The costs of the system components fabricated for this study included US$76 for each sensor node and US$55 for the gateway, totaling US$967 for the 12-node study described.