Optimization through Precision
Skycision Web App | UX Design | 3 Weeks
Challenge
Skycision provides farmers with actionable insights from drone-based aerial imaging by using computer vision and machine learning techniques. The company needed a framework for a web-based client interface that collected user data to feed their machine learning algorithm and provided flexibility for future enhancements.
Solution
In February of 2016, with a team of 2 other designers, we designed the framework for a web-based application. The web application fits into farmers' workflow and provides actionable insights through dynamic mapping, smart threat notification, intuitive flight planning, easy on-boarding, and non-intrusive data collection. Also, the web app's modular based framework allows for future feature additions.
Decoding the User's Mental Model
Prior to sketching or wireframing, we spoke with precision agricultural experts, farm managers, geological imaging experts, agronomists and viticulturists. We wanted a deep understanding of how farmers manage their farms, how they use and view aerial imaging, and how Skycision's tools fit into their workflow.
A TOOL, NOT A SERVICE
Aerial imaging will never completely replace "boots on the ground": When alerted of a problem through Skycision's client portal, farmers still need to investigate the specific area, confirm the alert, gather samples, and provide some analysis. Therefore, Skycision's client portal was designed as a complementary tool to accelerate problem identification, rather than something that replaces their current farm management system.
Information architecture
We determined a modular based layout with a default map view was the most intuitive because Skycision’s tools provide location based insights.
MAPPING INSIGHTS
Current methods of aerial image-based analysis requires more technical capability than is typical among farmers. Users thought NDVI maps were "cool," although they were often unable to drive actionable insights.
Some viticulturists in Napa received monthly NDVI maps from aerial imaging companies that used planes instead of drones. However, the images were often not calibrated which resulted in unreliable data.
Dynamic, Explorable Mapping
Comparable, dynamic maps tell a fuller story. Users can explore multiple maps side by side or by using the slider tool. In addition, users have the ability to import and compare their own maps (GIS or Shape files).
Farm activities are subject to many phenological variables and based on events - planting, pruning, harvesting. A timelapse feature allows users to view how their crops progressed compared to previous years.
ALERT INSIGHTS
Users care about optimizing inputs and quickly identifying threats to maximize the quality and quantity of their crops.
Accelerated Threat Detection
Skycision's computer vision technology spots inconsistencies and alerts the farmer. After scouting the area, farmers have the ability to classify the problem area, then prioritize and track it; which in turn provides data to train Skycision's machine learning algorithm.
Flight Insights
Similar to a Doctor diagnosing Cancer, prescribing medicine, and scheduling routine CAT scans; Farmers identify crop problems, prescribe solutions, and monitor it by severity. Skycision's aerial imaging and NDVI mapping is similar to a cancer patient’s CAT scan.
FLIGHTS AS PRESCRIBED
Farmers are familiar with the shape and location of their fields, therefore, within the "Flights" module, flight selection is location based. Farmers select a field(s) from a map view and schedule flight detail from an intuitive minimal sidebar. The bottom timeline panel displays a preview of past and future flights on the selected field to help prevent double booking.
final thoughts
We learned late into our design process that Skycision was partnering with a farm management company, Renteria Vineyard Management. Renteria would sell Skycision's tools as an addition to their existing farm management service. Skycision's shifting business model meant the web app interface would need to cater to different users with multiple permission levels and/or features, which was not considered in the web app's design.
Our presentation was well received and I was asked to continue with my user research as well as design the framework for Skycision's mobile app.