Soil Sample Management
One of the biggest challenges many businesses face is finding data management software that fits their needs. There can be a number of reasons for this, ranging from working in a small or highly specialized industry, to having a complicated internal workflow. For these situations, a custom software and database solution can fit the bill.
Platte Valley Labs is an agricultural, enviromental, and wastewater lab in Gibbon, Nebraska. As one can imagine, lab work can generate quite a bit of data. We were approached by them to assess their agricultural soil sample storage and management systems.
For those unfamiliar, agricultural soil sampling is the process of analyzing the micro and macro nutrients of soil from one or many locations. This data is compiled into a report with recommendations for fertilizer and nutrient application. In more advanced setups, soil samples are taken in a grid across the entirety of a field and analyzed, creating a nutrient topography. With modern farming techniques, fertilizer and nutrients can be selectively applied in only the locations needing help and in the quantities necessary.
This process proves quite practical for ag producers, as even a small amount of investment in soil sampling can save a sizable portion of both money and time.
The old system being used to manage the soil samples utilized a combination of a Paradox database and Excel workbooks. The general work flow involded passing around Excel workbooks containing different lab results. Rows and columns were then copied and pasted between them. Some of this process was simplified through the use of macro keys, though this only sped up the workflow marginally. After agrigating all the results, the final workbook was reviewed before being stored in the database. Reports were then loaded from the database and printed. There were 3-4 different templates, depending on the sample submission type, that could be printed.
One of the major challenges when starting a new project is becoming familiar with the industry. This is known as acquiring domain knowledge. Everything from the vocabulary used to common practices can be a challege to learn. The number one tool in acquiring domain knowledge is communication. Often, the questions you most need answered are the ones you'd never think to ask. These questions are usually about knowledge that is so obvious to the client, they didn't think it needed mentioning. In this case, I missed one particular detail when creating a rough draft of the software: the vast majority of the time, soil samples are sent in by agronomists, not by the ag producer. Agronomists contract with producers to help with soil and crop management.
PVL's customers are the agronomists. Each agronomist has one or more producers, and each producer has one or more fields. While an agronomist may submit samples from many producers at once, they are broken into individual submissions on a per-producer basis. This gives the agronomist a seperate report for each producer.
The process starts with an agronomist dropping off a sample package at the lab. This will be a mixture of samples from many producers, accompanied by a sample submission form. This form includes the agronomists designation for each sample, the producer it came from, the field it came from, and any crops the producers is considering growing.
Each sample package is broken down into submissions and recorded. From here, the samples are given a lab number. This is how each sample is tracked through the analysis process.
The lab can only process so many samples at once. A group of samples processed at the same time is called a batch. A batch of samples are then placed into numbered storage trays, ranging from 1 to 20. These trays move from station to station. At each station, a small scoop of soil is taken for that analysis. Each analysis produces an output that is recorded under that sample's lab number. After the lab work is complete, what remains of the samples is discarded.
After the raw data is collected, various metrics are calculated and recorded for each sample. Some of the metrics are fairly straight forward, such as Sum of Cations, while others are quite intricate. For example, if the producer is considering growing corn and wants to achieve a yield of 250 bushels per acre, the report will detail how many pounds of nitrogen to apply per acre to reach the target yield.
Finally, the raw data and calculated metrics are both compiled into a single report that is given back to the agronomist.
The software we developed was an entirely web-based platform. All data entry, importing, and exporting is done through the browser. Agronomists are also given a permission-based login to access their data and print reports and documents.
From the beginning of the process, the soil sample submission form is available to agronomists online. This form includes their contact information automatically. They can then print these forms on their own printer. From the start, this makes the process all the more efficient by reducing the need for PVL to keep pre-printed reports on hand, eliminating the need for agronomists to enter their contact information on each report, and shortening the delay in sending and receiving these forms.
Logging in soil samples requires an incredible amount of data entry and is a very time intensive process. Because of this, the user interface for data entry needed to be fast. We created a layout that seemless moves from submission to submission, sample by sample. To further streamline the workflow, this entire layout can be easily controlled and navigated via the keyboard alone.
As samples are batched and processed through the lab, each lab tech needs to easily locate every lab number's sample. In addition to the new software creating the batches, it produces a color coded Excel report of where each sample can be found.
Some lab equipment is capable of directly exporting data to a file. The software provides the ability to import this data directly from the output format.
After all the lab analyses are complete and the results are added to the software, a report is generated with both raw and calculated results. This report has been streamlined to present all necessary data in a legible, easily navigated layout.
Depending on the agronomist, the report will then be either mailed or made available online via the agronomist's account.
Custom software and databases are often the most effective tools for highly-specialized industries. They can be tailored to fit the current company workflow, while still saving time and money.
If you are interested in a similar solution, please reach out.