Computational models of landscape tree markets
Forecasting the prices and distribution volumes of landscape trees from transaction, cultivation, and weather data.
Independent non-profit research group · Seoul
Computing living landscapes.
An independent, non-profit research group formed by members and alumni of Seoul National University. We treat landscapes as measurable systems, from tree markets and container cultivation to the traces people leave on their own screens.
Read our researchWhat we do
Alpha Lab studies landscapes as systems that can be measured, modeled, and forecast within honest limits. Our first line of work grew out of the landscape tree industry: predicting prices and distribution volumes, and running the production of container-grown trees on data rather than habit. A second line turns the same discipline toward people, making a person’s own digital activity searchable as a kind of external memory.
We borrow our standards from physics. Quantities are measured before they are modeled, models are written down explicitly, and every claim carries its error bars. We are small, independent, and non-profit, and we publish what we learn as notes on this site.
Research
Programs grounded in registered and published patent work.
Growth data and production systems for container-grown trees
An ERP-style system that records growth and environment data for container-grown landscape trees and recommends production conditions.
Personal activity search and memory interfaces
Making a person's own desktop and browser activity searchable, as a form of external memory.
Suchan An 안수찬
Seoul National University (2012–), Landscape Architecture and Computer Science and Engineering, double major
Suchan An founded Alpha Lab and leads it as Principal Investigator. Trained in both landscape architecture and computer science at Seoul National University, he works on the boundary between the two: measuring living landscapes and building the software that reasons about them. He is a co-inventor of two registered Korean patents on machine-learning prediction of landscape tree markets and on production systems for container-grown trees (2021), and the applicant and sole inventor of a 2025 application on desktop and browser activity search, published in 2026.
Notes
Short explainers on the ideas behind our work.
Searching your own screen
A search index of what one person has seen is technically tractable. The hard part is that it would be the most complete record of a person ever built.
What a container tree ERP actually records
Growth, environment, and handling, measured on a schedule. Once every tree carries its history, the database becomes the industry's missing standard.
Why landscape tree prices are hard to predict
Lag, lumpy demand, and private quotes make the landscape tree market resist forecasting. The honest response is to measure it anyway and report the uncertainty.