Skip to content

RFID Moisture

Jump to bottom
mattguo5 edited this page May 17, 2021 · 45 revisions

RFID Moisture | Updates | Github

Project Team:

Electrical & Firmware Lead: Matt Guo and Natalie Clouse

Mechanical Lead: Max Chu

Overview

OPEnS RFID aims to address common overwatering and financially demanding issues to agricultural grown by providing a cheap, wireless, battery-less solution to large area moisture sensing. OPEnS RFID integrates the newest advances in ultra-high frequency (UHF) radio frequency identification (RFID) technology into our sensors, which make them a fraction of the cost of other sensors. When set up on an irrigation boom at a common greenhouse, all moisture processing can be done locally onboard a powerful microcontroller, allowing for precise and autonomous control over the watering of these crops.

Current Problem

Overwatering is ubiquitous in agriculture. Whole fields are grossly overwatered to ensure that no single plant experiences water stress. Farmers typically overwater their crops by 20-50% as a precautionary measure. Higher yields and low water costs make this behavior rational. However, recent improvements in water sensors and precision irrigation will make it possible to significantly reduce agricultural water use while maintaining yields.

According to the World Wildlife Fund, agriculture consumes 70% of the planet’s accessible freshwater1. For comparison, 8% of water is used municipally for drinking, sewage and industry. In water strained regions like California, Northern China, and Pakistan, increasing agricultural and municipal demands compete for a finite waters supply.

Another impact of overwatering is agricultural runoff. When soil is oversaturated, the resulting runoff carries environmentally hazardous pollutants and fertilizer into local water systems. Fertilizers consist of nitrogen-based compounds that are harmful to freshwater ecosystems. This runoff also depletes soil of its nutrients, further exacerbating the need for fertilizer.

Precision irrigation systems exist and can substantially reduce agricultural water use; the current bottleneck for these systems is soil moisture data. Without corresponding moisture data the water savings for these systems largely go unrealized. The soil moisture sensors that exist are either too costly, obstructive, or too inaccurate to use in conjunction with precision irrigation. In the following section we describe these limitations.

Hardware Description

The main components of this system are the Passive RFID Tags, RFID reader, Adafruit Feather M0, and a relay. Passive RFID tags are inserted into the soil of a potted plant the rest of the components are housed in a pelican case that is attached to an irrigation boom with the antenna also attached to the boom. Inside the pelican case the relay, feather M0 LoRa, hypnos board, and RFID reader are stacked on top of each other. The feather M0 gets power via a 5V 2A AC to DC wall power adapter through micro USB the feather then supplies the hypnos and the relay with 3.3V and the RFID reader with 5V.

Block Diagram

Current Solutions

The two primary solutions for soil moisture data have been capacitance probes and satellite imaging. Capacitance probes are the industry standard for taking cheap, accurate soil moisture measurements; however, capacitance probes require a continuous supply of power and often need to be checked manually. These sensors are relatively cheap but installing the number of probes necessary for a precision irrigation grid is costly. For these reasons capacitance probes are impractical for large area moisture sensing.

Satellite imaging is a cost-effective solution, but the results from these images come sporadically (less than once per day) and have low granularity (measure on a hectare scale). For these reasons they are ineffective for managing water usage on a day-to-day basis.

Outcomes

Initial Adafruit STEMMA Soil Moisture Sensor Trial

Initial field testings were done with the Adafruit Capacitive I2C Soil Moisture sensors at Peoria Gardens, mapped to their subjective 1-5 rating scale, where 1 is considered very dry and 5 is considered very wet. The Adafruit Capacitive Soil Moisture sensors contain 10-bits of resolution, designed to serve as the control for the RFID Soil Moisture sensor moisture correlation. The idea behind these data points is to provide accurate mapping of the RFID sensor soil moisture to the proven Adafruit Capacitive I2C Soil Moisture sensor. Updates on these correlations can be found in the Updates page.

Proven Correlation with the 5TM Moisture Sensor

After conducting 48 trials at 9 moisture levels, the data showed a correlation between the RFID tag's sensor values and the 5TM's measure of relative permittivity. This implies that SmarTrac's Dogbone RFID tags are sensitive enough to be used as a moisture sensor. Using the linear, empirically derived equation to map the RFID values to the 5TM,g(x), and then the Topp  equation (Topp et al 1980) to map from relative permittivity to volumetric water content (VWC),f(x), the equation relating the RFID to VWC is below:

f(g(x)) = -8.82803×10^-8x^3 -2.4733710^-5x^2 -3.8844710^-3*x+0.306477

Field Testing

After the correlation was proven in the lab, the team traveled up to OSU's Hermiston Agriculture Research Center to conduct field tests. Using a portable version of the RFID reader that sent data to a companion app over Bluetooth, we read the RFID tags and moisture values as a center pivot irrigation arm passed over. Using this app, we were able to visualize the water seeping down to the depth of the tag.

Future

Additional Projects

The correlation between the SmartTrac Dogbone RFID tags to soil moisture sensor can lead to many different projects for the OPEnS Lab. Currently, for RFID Irrigation System, soil moisture data is done locally onto the system via an onboard microSD card, running autonomously. On an Irrigation System with cellular connectivity, a TCP connection can be made on the system, streaming moisture value data onto a Google spreadsheet rather than a local .csv file. This can lead to a more manual control on the irrigation boom as well. With internet connectivity, a GUI can be implemented to adjust different calibration curves of the RFID tags, thus changing the rate of plant watering.

Resource List

Tutorials

Keywords

  • RFID

  • Dogbone

  • Soil Sensor

References

Home

Publications

Project Progression Guide

Projects

Resources

Wiki editing

Clone this wiki locally