PINGXIN WANG

LFR examines how physical processes become measurements, records and digital claims — and whether the relationships between them still hold.

A number can be calculated correctly and still refer to the wrong object, time interval or process. A record can be authentic and a model reproducible while the relationship required by a claim remains unestablished.

Our current work traces these transitions by preserving what arrived, examining the bridges between records and claims — including identity, time, units, measurement basis and missingness — making transformations and assumptions visible, and distinguishing what the available evidence supports from what remains unresolved.

Current research includes Battery/BESS and the GB power system, where we are developing bounded, traceable observation records and analytical tools that preserve member, temporal and evidential relationships where they matter.

These are current research settings, not the outer boundary of the work. Similar questions arise wherever measurements, simulations, operational records or institutional data are used to understand physical processes or support consequential decisions.

We are also exploring how these principles apply to human–AI analysis: whether the conditions supporting a comparison or interpretation remain active through data selection, calculation, visualisation and reporting, and whether new evidence can change affected outputs without discarding findings that remain supported.

LFR does not treat uncertainty as something to remove through confident language. We want to locate where uncertainty begins, what remains observable, what additional observation could distinguish competing possibilities, and where a claim should stop.

What does your digital record actually represent?