GridReady
Finding land is only the first step in a data-centre project.
At the Hong Kong AI Summit hackathon, our team took on CLP’s Urbanization brief. We narrowed a broad energy challenge to a concrete decision: a data-centre site may be secured, but can the grid actually serve it? GridReady lets visitors place a proposed load on a Hong Kong map, review simulated grid stress over 48 half-hour steps, and compare potential fixes. We built and pitched the prototype in 36 hours and won the hackathon.
Choose a site and power demand.
See simulated overloads across the day.
Compare upgrades and alternatives.
From the question to the solution.
I helped identify grid feasibility as the problem our team would tackle and played a substantial part in shaping the solution we proposed. That meant turning CLP’s broad brief into a question people could test, then finding a clear way to present the answer. I also helped design the GridReady website, which explains the prototype and the reasoning behind it.
Explore GridReady Read Artem Kuular’s project write-up
Take a closer look +
The question
How can an early site decision account for grid constraints?
The approach
Visitors place a proposed data centre on the map, review the simulated grid over 48 half-hour intervals, then compare fixes. The public-data model spans 90 nodes and 151 lines; some network layers and equipment ratings are estimates.
What it puts on the table
Capacity constraints, overloaded equipment and upgrade options with estimated cost and duration. The goal is to make an early feasibility conversation more concrete.
The result
A working team prototype, a 90-second video submission, a three-minute live pitch and a hackathon win in Hong Kong, October 2026.
The boundary
The model uses public data and estimated ratings. Its outputs support an early feasibility conversation; they are not measured CLP capacity, an engineering assessment or a connection approval.
