LV Digital Twin Platform
Open analytics for the
invisible grid.
A shared research platform for LV distribution network analytics. Each year, a new cohort of students builds directly on the work of the year before — implementing new methods, extending existing tools, and connecting analyses so that outputs from one stream become inputs to the next.
Research
What we work on
We tackle a range of problems in LV distribution network analytics. A few areas of particular focus:
Topology Estimation
Who is connected where?
Inferring the physical connectivity structure of LV networks from smart meter data alone — which customers share conductors, how branches connect, and where network boundaries lie.
Impedance Estimation
How does current flow?
Estimating the electrical impedances of conductors in the network. Line impedances underpin power flow modelling, loss estimation, and cross-validation of topology inferences.
Virtual Sensing
What is the transformer doing?
Estimating transformer loading and operating state from customer smart meter readings taken downstream — without any dedicated hardware at the transformer itself.
Situation Analysis
What if things change?
Scenario exploration tools for network operators and planners: what happens to voltages and loading as solar uptake increases, EV adoption grows, or customer load patterns shift?
A unified analysis platform
The platform exists to make student research cumulative. When one group implements topology estimation, the next can feed those results into impedance estimation or phase grouping. A shared interface gives every student visibility into what's already been built — reducing repeated work and letting each cohort start where the last one left off.
Python · OpenDSS via opendssdirect · NetworkX · Streamlit
Rolling cohort, growing capability.
Each year's students build directly on the work of the year before. See who has contributed and what they built.