Planting Simulator

Empowering Farmers with Predictive Planting Insights

Enhancing Precision Agriculture with Data-Driven Planting

Duration:

2022 – 2024

Problem Statement

How might we leverage a first of it’s kind crop phenology model that uses decades of data and cutting-edge research to create an intuitive digital tool for optimizing planting decisions?

Product Overview

The Planting Simulator project aimed to revolutionize agricultural practices by leveraging advanced technology and data-driven insights. Our goal was to develop a digital tool that harnesses decades of historical data and cutting-edge research to predict optimal planting times and conditions for various crops. By aggregating data from diverse sources, the tool provides precise recommendations to farmers, empowering them to make informed decisions and maximize crop yield and efficiency. Through collaborative research and iterative design, we aimed to deliver a powerful yet user-friendly tool that would streamline planting processes and drive sustainable agricultural practices.

Responsibility

Lead Designer: Steered the project as the primary designer, overseeing all aspects from initiation to delivering several version releases with a global user base.
Designer Responsibilities: Managed strategy, UX architecture, and UX research. Ensured teams were informed of the product's purpose and progress.
Delivery Lead Responsibilities: Filled in for several Delivery Leads during the project's lifecycle, ensuring continuity and efficient project management.
Beta Release Management: Orchestrated the beta release, gathering valuable user feedback and facilitating ongoing research for iterative improvements.

Tools

Target Audience

Emphasizing the need for a tool that caters to diverse roles in the agricultural ecosystem.

Primary Users:

Agronomists
Field Operations Managers
Seed producers

Secondary Users:

Digital Agronomist
Crop advisors
Crop scientists

Change is Hard

Agricultural professionals are often hesitant to trust new tools like Planting Simulator due to concerns about its reliability and potential impact on yield and financial gains. Their skepticism about the accuracy of digital simulations and fear relying on unfamiliar technology, especially considering the significant financial risks associated with poor yield or crop failure is certainly valid. From a user stand point, this was the greatest hurdle to overcome.

Pain Points

From

One regional planting recommendation per product.

Recommendations based per country based on regional averages and ideal conditions.
Manual tracking of weather conditions by agronomist.

To

Field-Level recommendations based on local factors.

Recommendations for each field based on specific planting dates and field specific data like soil composition.
Weather predictions are part of the engine.

So what?

User focused and business optimized

Better nick = more saleable yield
Simplicity – straight answer to “what day should I plant?”
Accelerates results – standard process drives better communication & faster decision-making

Use Cases

Pre-Planting, During Planting, and post Planting

Constant assessments to ensure the best outcomes to maximized yield.

Impacts of estimated planting dates,

Investing the relationships of male pollen to female silk.

Visualizing of key pllen managment activiteis

When will field of the same material start to flower (5% - 50%)

Running Best case and worst case scenarios

When things don’t go as planned what is the next best option.

Tracking fields through the growing season

Actual planting dates override simulated dates, ensure all data is up to date and valuable.

Why it matters

Supply Planning

How much, where and when

Increased Yield

Female pollination period is about 4-5 days

Unique Features

The right people on site at the right time

Risk Management

Weather trends can cause catostrophic effects

Challenges Faced

During the development of Planting Simulator, several significant challenges were encountered, each requiring innovative solutions and a high level of adaptability.

Scope Creep

The project faced scope creep due to the diverse range of stakeholder ideas and expectations. Each stakeholder had their own vision for the tool, leading to a lack of clarity and a constantly expanding scope. To address this challenge, extensive workshops and Lean UX Canvas sessions were conducted to align stakeholders and narrow down the project scope to its core objectives.

Simplifying Complex Logic

The underlying algorithm powering Planting Simulator was highly complex, making it challenging to create a user-friendly interface. Simplifying this complex logic for a simplified UI while retaining its predictive power required careful consideration and collaboration between UX designers and data scientists. Regular whiteboarding sessions were held to visualize user flows and identify patterns that could be mapped into the UI.

Role Changes and Adaptability

Throughout the project, the team faced several key position changes, necessitating adaptability and flexibility. As a lead UX designer, I had to step in as a subject matter expert for all aspects of the tool, including complex algorithmic logic. This required providing guidance and support to team members who were navigating unfamiliar roles, ensuring continuity and progress despite personnel changes.

Understanding Model Complexities

The team encountered difficulties in grasping the complexities of the predictive model behind Planting Simulator. The intricacies of the model were not readily apparent to all team members, leading to confusion and inefficiencies in development. To address this challenge, I created decision trees, flow diagrams, and use case scenarios to help visualize the logic and facilitate better understanding among team members.

Overall, overcoming these challenges required a combination of effective communication, collaboration, and problem-solving skills. By addressing each challenge head-on and leveraging the expertise of the team, we were able to successfully navigate the complexities of the project and deliver a valuable tool for precision agriculture.

Information Architecture Journey

The creation of information architecture for Planting Simulator was a collaborative journey that involved stakeholders from diverse fields. Through extensive research and iterative testing, we uncovered the tool’s potential, engaging with experts to identify key user needs and business requirements. This approach allowed us to design an architecture that addressed immediate challenges and set the stage for future innovations in agriculture technology.

Its started with a vision of cohesion

Phonology and Seed Production Prediction Vision

The vision of unifying large sets of complex and chaotic data and to present a simple, actionable solution to users has been a dream of agronomist for a long time. Agronomy may be cyclical, but its most value trends span decades and encompass thousands of data points.

Understanding its potential through use cases

Pre-Planting, During Planting, and post Planting

Constant assessments to ensure the best outcomes to maximized yield.

Impacts of estimated planting dates,

Investing the relationships of male pollen to female silk.

Visualizing of key pllen managment activiteis

When will field of the same material start to flower (5% - 50%)

Running Best case and worst case scenarios

When things don’t go as planned what is the next best option.

Tracking fields through the growing season

Actual planting dates override simulated dates, ensure all data is up to date and valuable.

Breaking it down by identifying potential patterns

By conducting iterative discovery sessions with stakeholders, I identified some of the most common use case patterns and flows essential for realizing the product vision. These sessions provided foundational knowledge to serve as a baseline for workshops and future discovery seasons, optimizing the utilization of subject matter experts’ time.

Expanding on its potential through workshops

Workshop session proved to be highly valuable in this case with a wide variety of stakeholders present to help work through what was desired and what was possible.

Translating concepts into actionable flows

First concepts proved to be less than ideal, as they exposed potential flaws with the core logic of the simulator. I created a simplified version as a fall back for MVP that would work in all use cases, but add more task on the user. This simplified version helped our team get past trying to tackle potentially unsolvable problems due to roadblocks and focus on what could be done.

I broke down all the potential uses cases and their user patterns in a new way to help the team reset and work through the problem in a more linear visual way for further iteration on the logic flows, I also included those use cases that could break the simulators logic in order to expose the solutions. After initial review with core team members, the use case patterns where brought to a white boarding session to further discuss their viability with our subject matter experts.

The outcome was a validated decision tree outlining the logic and user flow that accommodated all identified scenarios.

Low Fidelity Ideation

Much of my ideation is tracked through handwritten notes that eventually deliver low fidelity concepts. This is one example of how that comes to life. One core reason for this approach of written ideations over drawings is that I established design patterns early on in Blueprints lifecycle that would allow the product to grow in a structured and organic form. This means that it can actually be more time consuming to draw each flow idea since most elements have a natural place in the product. I found, for Blueprint at least, writing the hierarchical order of informations was more affective than drawing out the same concept with tiny iterations each time.

High Fidelity Screens

The seemingly simple interface is built around a strong information hierarchy and is back by user feedback from continuous communications with user. By keeping the users on a single e screen for most of their time using Planting Simulator they found they were able to stay on track and accomplish their goals in less time.

Key Takeaways

Collaborative workshops with stakeholders and iterative discovery sessions were instrumental in identifying common use case patterns and flows.
Valuable insights from workshops with diverse stakeholders helped clarify desired outcomes and feasibility.
The use of continuous feedback loops and iterative testing was instrumental in refining the simulator’s functionality and usability.
Planting Simulator’s predictive capabilities, intuitive interface, and flexibility have enhanced decision-making processes for agronomists and stakeholders in all regions.
The project showcased the value of cross-functional collaboration and user-centric design in developing complex agricultural tools.

Results

Successfully integrated into Blueprint.
Increased adoption and user satisfaction, with full adoption in North America.
V2 launch in LATAM ahead of schedule after increased demand.
Enhanced V3 features driven by ongoing research.
Increased yields year-over-year.
Showcased at global agronomy conferences.
The team was recognized for excellence in product delivery.