Pine Beetle.
a web platform to enable visualization & prediction of southern pine beetle infestations
PROBLEM
Background
Southern Pine Beetles are an extremely destructive pest of pine from the Southern United states.
Each year, they cause extensive economic and ecological loss to all types of forests across the country. For example, due to pine beetles, Alabama loses an average of:
60,000
pine trees / year
$800,000
of economy / year
Team: 3 developers, 1 project manager, 1 designer & me
Duration: 10 weeks (jan-mar 2022)
Skills: User research, brainstorming, user-testing, user-interface
Role: I collaborated with a fellow designer at Dartmouth's Applied Learning and Innovation Lab to re-design the user flow of two main website pages for the forest management center.
Impact:
-
Increased core task completion rate by 38%
-
Uncovered a new pain point and launched a new probability confidence feature
With the devasting impact and unpredictability of pine beetles, it is difficult for forest rangers and forest managers to decide how to best allocate resources. Over-allocating means that a large amount of resources and time are wasted and under-allocating leads to unsuccessful prevention of pine beetle infestations.
OPPORTUNITY
How might we help forest rangers & forest managers understand the past behavior of pine beetles and decide how to best allocate each year's resources?
SOLUTION
PROCESS
OVERVIEW
Project
Goals
With Pine Beetle being an established platform for multiple years prior, our first step was to conduct user testing on the existing platform to understand the current user experience and identify areas of improvement. Through interviewing multiple users, we found two main areas of improvement to focus on:
Paint point #1: There are varied discrepancies between each year’s predictions and outcome and it is difficult to judge the accuracy of the current year’s predictions.
Solution: Adding a predictions vs outcomes feature
Pain point #2: It is difficult to find the right information at certain points in the current user interface - such as how to toggle between counties and federal land, how to obtain information through the hover system and how to download data.
Solution: Improving the current user flow for obtaining data
RESEARCH
User
Personas
Forest Rangers
(major)
-
collects pine beetle data
-
uses data to prepare for infestations
-
minimally tech-savy
-
old age
Forest Managers (major)
-
uses data to manage resources at the top level
-
minimally tech-savy
-
old age
Researchers/scholars (minor)
-
uses data to explore biological areas of interest
-
minimal background knowledge of pine beetles and methods
-
young and old ages
RESEARCH
User
Research
Major Insights
We found that users were mostly happy with the platform, but wanted a capability to view the accuracy of past predications as well as finding a few confusion points in the user flow (mapped below):
IDEATION
&
TESTING
Iterations
Ideating a more intuitive user flow:
Using the mapped out user insights, we collaborated to design multiple iterations of possible user flows, regularly user testing each of them.
Multiple iterations of user testing and re-designing led to the final user flow, which consists of an intuitive flow (outlined below):
8/8 users were able to find the information they needed, compared to 5/8 initially
Download Data
I independently re-designed the Download Data popup to eliminate confusion points, which was validated through user testing
DEFINE
Specifications
Predictions vs Outcomes tool
To create a tool that met users’ needs, we used insights from user testing to create guidelines for the feature:
1. Have separate designs for two different views – since users found both graph and map view to be essential
2. Have the tool be as consistent with current website as possible – since users aren’t very tech savy and find it challenging to learn new functions
IDEATION &
TESTING
Iterations
Graph view design:
-
Since the predict outbreak page did not yet have a graph view, we wanted to ensure consistency through creating similar design to the Historical Data page graph view.
-
User testing: Users found this new graph feature comparing predictions to historical outcomes to be clear and very helpful
Map view design:
Since a map view already exists on the Predict Outbreak page, I ideated two ways to incorporate the new feature.
-
Design 1: Add a third circle (blue) to the bottom left corner to show divergence.
-
Pros: simple, fewer toggles
-
Cons: provides less detailed information
-
​
-
Design 2: Create a new toggle (top left corner) to a new map that displays divergence between predictions vs outcomes.
-
Pros: detailed & varied information
-
Cons: more implementation for devs, more complicated user flow​
-
