Helping horticulture farmers to analyse farm data and take better decisions — A UX and UI case study.

Published in

Fasal

10 min readMay 22, 2021

Introduction

Farming in India is dependent on a lot of guesswork and climate. It is largely a traditionalist ecosystem where the knowledge of farming is passed down from one generation to another. A son will most likely adopt the same farming practices and cultivate the same crop, which his father has cultivated for years even if it is not a very profitable crop. Now, this cycle is breaking with the spread of the internet and education. Farmers are progressing and are open to trying new technologies to improve their profitability.

About Fasal

At Fasal we help farmers predict ideal growth conditions, resource requirements for their crops, and alert them from potential pest/disease attacks by democratising frontier technologies. We use machine learning to predict rainfall at the farm level with high accuracy by collecting data from our sensors in the field along with other large data sets.

Fasal collects various data points like micro and macro climatic conditions (canopy temperature & humidity, rainfall, air pressure, wind speed, etc), soil parameters, crop stage, crop growth characteristics through a sensor unit installed on the farmer’s plots.

This data is analysed and processed by Fasal algorithms to provide actionable intelligence on:

Accurate irrigation timings with the ideal application amount.
Early prediction of pest and disease problems
Accurate spray timings for maximum utility of sprays
Plot specific weather data, like rainfall alerts, wind speed, temperature, humidity etc.

The farmer receives the above information through the Fasal application on his mobile phone. My job here is to design this application in such a way that farmers can understand the information provided to them and make informed decisions.

Fasal app overview

After the farm IoT device is installed at the Farmer’s plot, the farmer has to integrate the Fasal application with the sensor unit. (A plot is a part of the farm where the farmer is growing a particular crop.) Farmer can do it while onboarding on the Fasal application. After that, the farmer needs to upload all the relevant information about his farm and the crops he cultivates (soil type, crop name and variety, transplantation date) and create a plot on the Fasal app.

Soon after the integration, the Farmer will start receiving the temperature, humidity and soil moisture of his plot on the home page of the Fasal app. Farmers can create multiple plots on the Fasal app. Upon selecting a plot on the Fasal app, the farmer can see more than 13 parameters about that plot, shown on the plot-details page. My objective of this project was to redesign this plot details page.

Project brief

To create a user interface for the sensor data which is easily understandable to our farmers in terms of how any parameter affects their crop, and what exactly they should do to improve profitability.

My Role

Being the loudest voice on usability and UI internally

Challenges

Farmers should be able to access any required sensor data with the least number of clicks.
How to make sure that the farmers can understand what each parameter means and how it affects their crop?
What exact action do they need to take after receiving the information about their crop?
How to help the farmer gauge the status of each parameter according to their crop and climatic condition?
How to help farmers compare the forecast of various data points
How farmers can see the trend of certain sensor data and make better decisions?
How to accommodate other regional languages with different scripts like Hindi, Tamil, Kannada, Marathi and Gujarati in the UI design?

Issues

The data in the UI is not clearly visible to farmers due to the small font size.
Difficult to view past data of each sensor. Need to keep switching between current and past data.
Relation of each parameter with crop growth
Improvement in the overall visual design consistency.

Research agenda

To gauge the current understanding of farmers about various parameters.
What is the prioritisation of each parameter regarding crop and crop stage?
What problems farmers are facing with reading the data in version zero?
Is the farmer able to navigate through the app easily?
Is the farmer able to quickly understand what needs to be done?

What did we learn?

We conducted interviews with many farmers based on version zero of the Fasal app and we observed;

The font in the previous Fasal app was very small and hard to read especially under direct sunlight.
Farmers usually compare the hourly weather forecast data to plan their activity for the day like irrigation, spraying etc.
Farmers have a busy schedule, so the accessibility of the important features is important to them. They prefer most feature should be available within one or two-click after opening the app.
Most digital apps which farmer used are Facebook, WhatsApp and YouTube.
Visual representation of the data would help farmers analyse their farm condition, especially those from a lower educational background.
Farmers mostly use the application in the morning to plan the day’s activity or at night to plan the next day’s work.
Farmers use the past trends of various parameters like temperature, humidity, leaf wetness, soil moisture to plan what will be the best time to irrigate so that crops will absorb the most water and fertiliser.
All farmers see the predictions regarding disease or pest attack and weather predictions.
Weather data is the most important data point for the farmer on the plot details page.
After the weather, irrigation is the most important thing for crops and farmers actively used Fasal to manage it better.
In the current application, it is difficult to access the trends of each parameter because of navigation.
The temperature difference between day and night affect the health/quality of the crops. Farmers need to track the difference and spray immunity booster to maintain the good health of the crop. High temperature in the day and very low temperature at the night makes the plants very sick, they become incapable of absorbing nutrients from the soil.

Page Structure

Iteration 1: How did we begin?

At first, when I joined Fasal, I discussed with my team, what is the goal of the user when he/she opens the plot details page? What are the current difficulties farmers are facing in the present UI? We analysed the usability issues of the present app using heuristics evaluation and tried to come up with a design that addresses these issues and also any past feedback we got from our farmers.

At that time the world shut down because of the COVID-19 pandemic. So, we couldn’t communicate with the farmers the way we wanted to. The idea was not to waste time and create something based on our current understanding and take feedback from farmers and iterate to come up with a better solution.

Iteration 2

We also wanted to explore another idea where we decided to categorize the sensor data into three segments: weather, disease & pest, and irrigation. The reason was to reduce the scrolling done by the user to read a certain parameter.

Another reason was that since we were loading all 15 sensor data parameters on one page in version zero, there was a chance where it would increase the loading time of the page. Some of our users are located in a certain area where internet connectivity is rather slow, so we thought that it would take a lot of time to load all 15 sensor data on one page with slow internet.

We also decided to integrate the disease and pest analytics with the plot details. In disease & pest analytics, farmers can see the risk status of the different disease or pest models (A model is an algorithm that tracks the environment for a specific disease or pest pressure) in their plot.

Iteration 3

After iteration 2, we decided to test UI with our farmers and discuss it with some agri-research team and sales team members.

What we found out was surprising:

Farmers do not like to go too much deep inside the app to get certain data.
Scrolling was not a problem for most farmers.
Farmers wanted the fonts to be even bigger because most of them use the app under direct sunlight.
Rather than actual numerical data, farmers wanted some visual context about how it affects their crop.
Not all farmers use the application in the same way. Based on the crop they were growing, priorities for each farmer is different based on which they track certain parameters in the app.

We continued digging deeper into the farmer's psyche. We found out that the farmer had an affinity towards the visual representation of the data more than numbers, but many of our farmers use the numeric data to track the environment of their plot, as well. So, we had to accommodate both ways of representing the farm data to the farmers.

So, we were back at the drawing board. Following are some of the things we did to improve the UI:

We visualised the parameters making relations to crop and crop stage upfront.
To increase the font size of current weather data and weather forecast, I needed to change the format about how the data is shown presently.
We also removed the categorisation of the data since it increased the number of clicks to get to a certain parameter.

Iteration 4

Now iteration 3 was like a wireframe, just to represent the idea. But now I needed to improve the visual representation of the data. Also, some of the feedback which I got from my colleagues were:

The fulcrum of the tracker is not visible clearly, and it will be difficult for the farmers to see under direct sunlight.
Inconsistency in the icons and improvement in visual design.
Various other versions of our farm IoT have fewer individual sensors as compared to our full-fledged ones. When there are fewer sensors, the design will be very inconsistent with the present UI with all 15 sensors.

To make the icons consistent throughout the whole app, I made an icon library.

To solve the above-listed problems I came up with the following solutions:

Some other options of different visual elements to show the status of a given parameter
Improved visual design.

Final Design

Option 1 was rejected because the fonts were not big enough. And the fulcrum in the tracker is too small to see clearly for the farmers.

Option 2: This option was reasonably preferred by many farmers, but the right tick mark indicator doesn’t explain where the sensor reading stands on the “safe to danger” scale. And also the differentiation between the two sensor readings could be more substantial.

Option 3: Farmers also had a moderate affinity towards option 3, but it has the same problem as option 2 about the sensor reading indicator. But cards design differentiates between the sensor readings better.

Option 4 was also rejected because of the sensor reading indicator.

The weather forecast design format was also not liked by the farmers because it was difficult to compare data between each hour or days in this format. It is better if the data for each day or hour is side by side.

After considering all the feedback, I came up with the final design.

Designed a new tracker which is clearly visible
Explanation about what each parameter means is provided upon clicking on any parameter card.
Weather forecast data is shown side-by-side, which makes it easy to compare.

We continue to take feedback from different farmers after we released the new UI and try to iterate more and more to solve every possible hurdle farmers face while cultivating their crop and make the app even more intuitive to use.

Every farmer is growing numerous crops of different variety in various soil types and climatic conditions all over India. So the number of use cases for using Fasal is huge in our country. Our constant endeavour is to accommodate most use cases in our design and address every little problem in our scope that farmers face throughout their journey of cultivating a crop.

Conclusion

Technology has penetrated every aspect of our lives, from socialising, food delivery, shopping, services etc. It's about time we use the power of tech to help our farmers grow more and better quality crops and improve their livelihood.

Using the Fasal system, farmers were able to save up to 30–50% of their irrigation water usage, reduce their spraying cost up to 25–30%, and increase their yield up to 15–25%.

For me as a designer, working at Fasal has been an absolute pleasure. It's a unique product whose users and goals are different from most conventional tech startups. Everything has been built from scratch, from excellent hardware platform to complex AI algorithms. Our goal is to build a tool that can help any novice farmer get all the knowledge required to get into horticulture and improve his livelihood. We are moving forward at warp speed towards that goal but we still have a long way to go.