Earthshot Labs · 2022–2023
A modern field tool for forest scientists — using AR and ML to make tree inventories 5–10× faster than conventional methods.
Data is essential to the conservation of the world's forests. Measuring trees is a tedious task done by scientists and field technicians to manage forest projects and estimate biomass and carbon sequestration.
I worked with a small R&D team of 3 engineers to research user needs, design and build the first working alpha, scope future features, and run a field test in Panama.
Earthshot Labs was awarded an NSF SBIR grant for this work.
How do you measure a forest?
Luckily, you don't have to measure every tree. A random selection of sample plots are taken. Scientists may use a range of methodologies, but a common practice is to define a plot by a center GPS location and radius.
Stuck in the past
While remote sensing investment has grown over the last decades, there has been little improvement in scientific field tools. Measurements are still collected with measuring tape, pencil, and notebook.
Gruelling work
In the field, scientists are juggling supplies while hiking down a rainforest ravine, avoiding fire ants — and then it starts raining. The work is gruelling and prone to inaccuracy.
Our initial field test showed that using the app is 5–10× faster than conventional methods. Furthermore, our app built trust with people on the ground, giving them a tangible innovation instead of just promises.
Measure tree diameter with computer vision
Alpha test build
Measure tree height with AR
This feature worked surprisingly well even for large trees. While there is a margin of error, it proved to be within the acceptable range compared to the traditional clinometer method.
We explored several UX flows, and the simplest version proved the most reliable.
Small tree
Large tree
Record forest sample plots using AR
While field testing, we were delighted to find the AR circle is perfectly accurate on a hillside — scientists usually do a bunch of trigonometry to compensate for the slant of the hill.
I worked on this prototype in Unity with our engineers and tested in a variety of natural terrain in Panama. The AR tracking worked even in tall grass.
Figma prototype
Unity prototype
Seeing the forest and the trees
Seeing all collected data alongside photos is powerful. Conservation projects generate loads of data that mostly sits in folders and spreadsheets.
I focused on displaying information very visually — making each screen layout distinct for easier browsing and giving a feeling of zooming in and out.
'Permanent monitoring plots', where the same plots are measured year after year, is a common technique. I used time as the primary organizational concept.
I structured the hierarchy as: Projects → Studies → Sites → Plots → Trees.
These are concept designs — not a ship commitment, but an artifact for the team to imagine what's possible.
Figma prototype
