Research & lab

Built on measurements, not assumptions.

The Core Model is developed against physical tests on real cells and modules. Every result we publish states what it was tested on and how it was checked.

Cells · physical testsModules · physical tests2 patent applications filed

Engineer connecting cell sense leads on a battery module in front of test-channel racks — Module instrumentation in the Dynovy test lab

What we test

Controlled conditions, many channels, real hardware.

Short tests are only useful if they are repeatable. We test under controlled temperature, sense every cell group, and run units in parallel under the same protocol.

Engineer placing a cell fixture on the test bench — Cell characterisationCapacity, open-circuit voltage and pulse tests on reference cells give the ground truth for each family.

Battery cells inside a temperature-controlled test chamber — Controlled temperatureClimatic chambers hold cells at set temperatures, so temperature effects are measured rather than assumed.

Top view of a battery module under test with temperature sensors taped to the cells — Modules, sensed per cell groupVoltage and temperature on every cell group show how the cells inside one module differ.

Cabling at the back of multi-channel battery test racks — Many channels in parallelBidirectional multi-channel test stations run many units under one protocol, so results compare fairly.

How we validate

Four rules for every result we publish.

These rules govern the blind validation on a first battery family, which is the next phase. The word validated is reserved for results that pass it.

Blind units

Accuracy is measured on units that were never used to calibrate the model.

Independent reference

Each estimate is compared with a full conventional capacity measurement on the same unit.

Metrics fixed in advance

Error, range coverage, correct grading, abstention and repeatability are defined before testing starts.

No result without evidence

When the evidence is not strong enough, the result is withheld and the unit referred to a full test.

Where we are

What exists today, and what comes next.

We separate what has been done on physical units from what is still being developed. Performance figures stay marked as targets until they are demonstrated.

Done

Core software implemented and verified on synthetic data
Physical tests on cells and modules, used for state-of-health evaluation and for developing the model
Two patent applications filed, in battery health assessment and in battery safety

Area	Next	Status
Cells	Current focusBlind validation on a first battery family is the next phase.	Current focus
Modules	In developmentPhysical tests at module level; the Core Model is being extended from cells to modules.	In development
Model inputs	In developmentBattery structure information, innovative triage methods and BMS data are being added alongside the short electrical test.	In development
Forecasting	In developmentDegradation and remaining-life forecasting under defined operating conditions.	In development
Packs	RoadmapPlanned after modules, with their own calibration and validation.	Roadmap
Test time and accuracy	TargetPublished once demonstrated on reference units; shown as targets until then.	Target

Intellectual property

Two patent applications have been filed.

They cover battery health assessment and battery safety. Method details are not published on this site.

Have reference data or units for a battery family?

A validation project compares Dynovy's short-test results with reference measurements on cells or modules you choose.

Discuss a validation project