The kilogram is no longer an object
Since 20 May 2019, the kilogram has been defined by fixing an exact value of Planck’s constant rather than by assigning one metal cylinder the world’s mass. Laboratories realise the unit through experiments such as Kibble balances or silicon-sphere measurements.
For more than a century, the International Prototype of the Kilogram—a platinum–iridium cylinder near Paris—was the definition. Copies could be compared with it, but any mass change in the prototype was conceptually impossible to detect as a change in the unit itself. The revised SI fixes h at exactly 6.62607015 × 10⁻³⁴ joule seconds, linking mass to electrical and frequency measurements through quantum standards. The definition does not prescribe one apparatus, so independent realisations can improve and compare. The old cylinder remains an exceptionally stable artefact, not the definition.
A unit can move from custody of one object to a reproducible relationship among constants, instruments and equations. That decentralises realisation and removes the last SI base unit tied to a unique artefact. It does not make measurement effortless: uncertainty now lives openly in the experiments that realise the definition.
Planck’s constant has no uncertainty in SI units because its numerical value defines part of the unit system. A Kibble balance still has alignment, gravity, electrical and mechanical uncertainties. Exactness belongs to the convention; uncertainty belongs to connecting a laboratory object to it. This separation is central to modern metrology and often misunderstood when people hear that a constant is fixed.
Kibble balances compare mechanical power with electrical power linked to quantum voltage and resistance standards. The silicon route counts atoms in an extraordinarily pure sphere to connect microscopic mass with a macroscopic object. Agreement between conceptually different realisations tests hidden biases. A unit embodied in a network of reproducible methods can be challenged in ways a unique defining artefact cannot.
Isn’t Planck’s constant just a more abstract artefact chosen by committee?
Its numerical value is a human convention within chosen units, but the physical relation can be realised independently wherever suitable equipment exists. A cylinder requires comparison with that one object and its chain of copies. The constant-based definition shifts dependence from custody and material stability to reproducible experiments and transparent uncertainty budgets.
A future SI revision could choose another equivalent set of defining constants, but measurements would be arranged to preserve continuity of the kilogram.