Here's the thing: it almost can. And the gap between "almost" and "fully" is one of the most consequential materials science problems on the planet, because concrete is the second most consumed substance on Earth after water. We pour about 14 billion cubic meters of it annually. And it all cracks.

The dominant approach is bacterial self-healing concrete, pioneered by Henk Jonkers at TU Delft. The idea: embed spores of Bacillus pseudofirmus or similar alkaliphilic bacteria into the concrete mix, encapsulated alongside calcium lactate as a food source. When a crack forms and water infiltrates, the bacteria wake up, metabolize the calcium lactate, and precipitate calcium carbonate — limestone — filling the crack from within.

Let's do the math on whether this actually scales.

A typical structural crack we'd want to heal is about 0.3 mm wide and penetrates perhaps 50 mm deep. For a 1-meter length of crack, the volume needing fill is:

V = 0.3mm × 50mm × 1000mm = 15,000 mm³ = 15 cm³

Calcium carbonate has a density of about 2.7 g/cm³, so we need roughly 40 grams of CaCO₃ per meter of crack. The metabolic pathway is:

Ca(C₃H₅O₃)₂ + 7O₂ → CaCO₃ + 5CO₂ + 5H₂O

One mole of calcium lactate (218 g) yields one mole of CaCO₃ (100 g). So to produce 40 g of limestone, we need about 87 g of calcium lactate. Current formulations embed roughly 0.5–1% bacteria + nutriite by weight of cement. For a typical beam with 400 kg/m³ cement content, that's 2–4 kg of healing agent per cubic meter of concrete.

Here's where reality gets interesting. Lab results show bacterial concrete can heal cracks up to 0.8 mm wide — impressive, since conventional autogenous healing (just the natural rehydration of unbound cement) maxes out at about 0.1–0.2 mm. But structural engineers care about cracks wider than 0.3 mm, because those are the ones that let chloride ions reach the rebar and trigger corrosion. The bacteria hit exactly the right range.

The cost problem: bacterial self-healing concrete currently adds about €30–50 per cubic meter to the roughly €80–100/m³ cost of standard concrete — a 30–50% premium. Sounds steep. But consider the lifecycle: corrosion-related repairs on a typical reinforced concrete bridge cost €200–400/m² of deck surface, and most bridges need major intervention within 30–40 years. A 500 m² bridge deck uses maybe 200 m³ of concrete. The healing agent adds €6,000–10,000 upfront. Deferring one major repair cycle saves €100,000–200,000. The payback ratio is roughly 10:1 to 20:1.

The real engineering bottleneck isn't biology — it's oxygen. The bacteria are aerobic. Deep inside a concrete member, dissolved oxygen runs out within millimeters of the crack face. Researchers at Cardiff and Bath are experimenting with shape-memory polymer tendons that contract when heated, physically closing cracks to bring the faces within bacterial healing range. Others are exploring vascular networks — thin glass capillaries embedded in the concrete that rupture on cracking and release healing agents, mimicking how blood vessels deliver clotting factors.

The vascular approach is especially wild: imagine a concrete bridge with an embedded circulatory system, periodically flushed with fresh healing agent from a reservoir. Ghent University has demonstrated this in lab-scale beams, achieving repeated healing of the same crack location — something bacterial methods struggle with, since the embedded nutrient supply is finite.

We're converging on a future where concrete isn't an inert block but a living composite — one that senses damage, responds, and repairs, all without human intervention.