PhosphoSense

Physical Design in Bioengineering

Individual

August 2023 - December 2023

Skills:

CAD (Fusion 360)
3D Printing/Rapid Prototyping
Experimental Design

User-Centered Design
Spectrophotometry
Sensors, Arduino, Embedded Systems

Overview

The PhosphoSense is a simple sustainable solution for farmers who would like to have an understanding of the level of nutrients within their soil. Generally, the 3 most important nutrients for soil testing are phosphorus, nitrogen, and potassium. Currently, the industry-standard for soil testing is to collect soil samples and send them off to a specialized soil lab. However, this often has a lag time of weeks, which means the results do not give current information on the nutrients in the soil. Having real-time information on nutrient levels within soil allows for farmers to use fertilizers in a more efficient and sustainable manner. Thus, there is a need for a real-time method to detect nutrient levels in soil. The PhosphoSense provides a simple, very low-cost method for the spectrophotometric quantification of phosphate levels in soil.

User Research

My task was to develop a device that could perform soil phosphorus tests on site. I interviewed with the Cornell Soil Testing Lab, as well as some farmers they put me in contact to figure out customer needs.

From the soil testing lab, I learned that measuring plant-available phosphorus in soil was not as simple as quantifying the total amount of phosphate within a soil sample, as much of the phosphate in soil can be immobilized by metal ions and thus made unavailable to plants. I also learned that the farmers I spoke to would use such a device if it was easy-to-use, cheap, and accurate.

Brainstorming

After I took everything into account, I researched other ways in-situ soil nutrient testing was performed. I compared these methods through a Pugh matrix to determine the best method I wanted to pursue.

After performing the Pugh analysis, I determined that the colorimetric method was the best for a few reasons. First, it allowed for the accurate detection of plant-available phosphorus, as many regional soil labs have already determined the best colorimetric way to determine plant-available phosphorus for their region. It would also be very cheap to produce, as ion-selective electrodes can be upwards of $1,000 if needed for use in a field setting.

Prototyping

The basic steps for my device were going to be:

Extract phosphorus from soil sample into solution chemically and develop color.
Measure phosphorus through spectrophotometry.

I wanted to simplify the process for the farmers as much as possible, so I found a commercially available soil-testing kit that used a simplified version of the Murphy-Riley process commonly used in soil testing labs to determine phosphorus concentration within solution. Thus, I could focus on the actual spectrophotometric device.

To prototype the spectrophotometric detector, I used the OPT101P ambient light sensor to measure the amount of transmitted light through a solution. I used an LED emitter that emitted at 880 nm to provide the light source, since I didn't need other wavelengths. I integrated an LCD screen, a potentiometer, and a push button into the device to allow for control of the device between different menus without a need to interface with a computer. I programmed an Arduino Uno to perform all of the calculations and control the device, and the entire device could easily be run with a 9V battery connected to the Arduino Uno. I designed the housing in Fusion 360 and 3D printed the device housing on a Bambu Labs X1 Carbon.

The control algorithm for the device was simple and follows the principles of a finite state machine. To measure absorbance easily, I first needed to measure the amount of transmitted light within a blank. I also needed to measure the absorbance of known concentrations to create a standard curve, from which I could then linearly interpolate new concentrations. I used the potentiometer to shift between different states and the button to perform the actions associated with each state. For the prototype, I assumed that the standard concentrations would be the same concentrations each time and would be the ones I used for testing.

Functional Decomposition

Diagram of control algorithm

Testing

I designed a variety of experiments to help validate and test the device. First, I validated the sensors were working and I could accurate measure the absorbance of the phosphate solutions. I then created a standard curve and tested the ability of the device to measure the concentrations of solutions along the standard curve. Finally, I tested the device on actual soil samples. I documented everything within my lab notebook and wrote a final report detailing the design.

Video of me demonstrating how to use PhosphoSense

Soil Sample

Standard curve