Start with the number that matters to a network that must run every morning: the Romans laid more than 400,000 kilometres of road, of which over 80,500 were stone-paved, and some of it still carries traffic. Not “still visible.” Still carrying. A design life that outran the state that commissioned it is the rarest specification in civil engineering, and it did not happen by accident.

The frontier here is not the empire — it is the cross-section. P. Andrew Karam, writing in Science and Its Times, describes the standard build-up: drainage ditches cut first, the spoil used to raise the roadbed at least three feet above the terrain, then layers of gravel, sand, concrete and paving stone stacked as much as 1.4 metres deep, the surface heavily crowned so rain ran to the side. The genius was not the stone on top. It was the water management underneath. Every century-scale asset I know of fails the same way: not crushed by load, but rotted by water that had nowhere to go.

←TODAY: 80,500 km of Roman stone-paving outlasted three signalling epochs and is still mapped in the Antonine Itinerary. →3012: the infrastructure that survives to the Zurich-3012 horizon will be the one whose drainage and diversionary route were drawn before the ribbon was cut. Fulcrum: durability is a maintenance decision made at design time, not a material property discovered later.

The lineage is worth naming precisely, because it is a governance story as much as an engineering one. At its peak the system ran to 372 great roads binding 113 provinces, a courier on horseback covering up to 121 km a day across them. The rules were written into the Lex XII Tabularum in 451 B.C. — a standard width of eight Roman feet on the straight, sixteen on the curve — before the Via Appia reached Capua in 312 B.C. Livy records the second-century-B.C. censors Quintus Fulvius Flaccus and Lucius Postumius Albinus as the first to award contracts to pave the city’s roads in stone. And the maintenance office came later still: as Jesús Rodríguez Morales records in National Geographic History, Augustus in 20 B.C. appointed the curatores viarum, magistrates whose entire job was awarding contracts and overseeing repair. The road-map, the build spec, and the named owner-of-maintenance — that is the full triad, and it is exactly the triad most modern megaprojects deliver two-thirds of.

On a working desk in 2026, this lands as a question, not a monument. When you detail a station forecourt, a viaduct deck, a transit platform — anything meant to stand for sixty years — the Roman cross-section asks: where does the water go, and who owns the ditch in year forty? PAZ has covered the failure mode from the digital side — our piece on the Swiss e-ID cut showed how the interoperability and back-up layers, the redundancy, were the first line-items deferred to save CHF 1.7 million. Physical infrastructure loses its second path the same quiet way. The single supplier, the drainage channel value-engineered out, the diversionary route closed “temporarily.” Then one storm-summer like the fires and heat now crossing France and Spain, reported this week by The New York Times, and the asset with no redundancy is a single point of failure wearing a public ribbon.

Atelier: The office move this week is to add one row to your asset schedule for every long-life element: drainage owner, year 40. Not the warranty — the maintenance owner. If the cell is blank, you have designed a Roman road without a curator, and it will rot from the ditch up. On Monday, open the last public-works detail you drew and mark, in one sitting, where the water leaves the structure and who clears it in decade four.

Hack: Size the crown that keeps water off my deck before you argue about the surface finish. The Roman road was crowned so rain ran to the ditch; the same fall governs a modern platform or bridge deck. Compute the centre-to-edge rise for Karam’s 4.6-metre trunk-road width at a spec drainage gradient — if it comes out too flat to shed a storm, the cross-section is wrong no matter how good the paving. Run this and read the millimetres:

width_m = 4.6        # Roman trunk carriageway (Karam, 15 ft)
crown_grade = 0.02   # 2% centre-to-edge fall, still spec today
rise_mm = (width_m / 2) * crown_grade * 1000
print(f"crown rise at centre: {rise_mm:.0f} mm")  # 46 mm

Forty-six millimetres of crown over a 4.6-metre width is the whole difference between a surface that sheds and one that ponds. The Romans knew it in stone; you spec it in a drainage layout. The number is small and the consequence is a millennium.

So the move is not to admire the Appian Way. It is to treat design life as a line-item you defend in the meeting where someone proposes to cut the ditch. Build the diversionary route, and name the curator, before you need either.

Source: europe.factsanddetails.com