The cost of optimising

On 23 March 2021, a container ship lodged itself across the Suez Canal and blocked for six days a corridor carrying roughly 12% of world trade. That same year, the semiconductor shortage cut global car production by 7.7 million vehicles, and the consultancy AlixPartners put the industry’s lost revenue at 210 billion dollars.

Both episodes share one feature. The affected systems were running excellently, at maximum efficiency, right up to the moment of disruption.

Stock as insurance

Just-in-time production, free of inventory, was one of the great industrial achievements of the last century. Capital tied up in warehouses falls, floor space is freed, cost per part comes down.

Inventory did perform a second function, paid for quietly: it absorbed shocks. Once removed, every upstream interruption travels instantly to the assembly line. Efficiency was bought with fragility, and the price stayed invisible until the first serious disruption.

Technical monoculture

Standardisation cuts costs. A single component type, a single supplier, a single protocol — all of them simplify design and cheapen production.

That same standardisation correlates risks. When a whole industry uses the same chip, one halted plant hits everyone at once. Agronomists have known the phenomenon for a long time under the name of monoculture; engineers are rediscovering it now, at the scale of supply chains.

Redundancy as a design brief

Here the inventor steps in, in a role rarely discussed. Resilience is designed, just as efficiency is.

Equipment conceived for graceful degradation keeps working at reduced capacity instead of stopping abruptly. A product conceived for a second supplier accepts components from different sources, with slightly different specifications. A device designed for repair lives longer than one sealed shut.

Each of these choices costs something in manufacturing and saves a great deal in the first crisis. They call for original technical solutions, and therefore patentable ones.

Industry 4.0, seen from this angle

Digitalising a factory opens a second front. An integrated system, with every machine connected, gains in visibility and loses in isolation. A network fault, a bad update or a cyberattack reaches the whole line at once.

Designing independent levels of operation — the line carrying on locally even with the central link cut — becomes a technical requirement, as serious as integration itself.

We therefore support an addition to the funding criteria for industrial digitalisation: alongside the productivity gain, the behaviour of the system under disruption. An indicator measuring how fast a line recovers after a stoppage says as much about its health as cost per part does.