PDOFoundation · feeds every engine

Power-system Data Orchestrator.

One pipeline. Every source. A single canonical model every other engine inherits.

The data pipeline that normalises every market and network feed, maps each economic asset onto its physical busbar, reduces a year of operation with scenario reduction, and hands a clean dataset to every Pillar-2 and Pillar-3 engine.

Multi-source ingestAsset → busbar mappingScenario reductionHPC multiprocessing
5Source platforms unified
→ ReducedScenario reduction
MinutesAnnual pipeline run-time
100 %Asset → busbar coverage
The signature output

Five feeds in. One canonical model out.

Heterogeneous market and network sources are normalised, reconciled to topology and reduced — then handed to every engine as one shared input layer.

Pipeline · sources → canonical model → engines · multiprocessing · days → minutes
New datasets, markets and tools plug in without rebuilding the foundation — every engine inherits the same validated, auditable dataset.
Inside the engine

The foundation, in five passes.

01

Normalise

Every source mapped to a Parquet schema with consistent units and conventions.

02

Node-merge

Topology reconciliation — duplicate and split nodes resolved against the network model.

03

Asset → busbar

Each economic asset bound to its physical node, so market and grid speak the same language.

04

Reduce & validate

A year of hours reduced to representative scenarios at <1% reconstruction error, then validated.

The read

PDO turns five incompatible feeds into one canonical model — so every engine downstream is reasoning over exactly the same, fully-auditable picture of the system.