The consensus on fusion is that it is always thirty years away, and the reliable way to lose money on it is to promise clean electricity and ask investors for patience. Astral Systems, a Bristol deep-tech company, is betting the other way. It has raised £23 million in a Series A first close led by Mercia Ventures, with Tees River, Daphni, Blast Club, Speedinvest and Playfair Capital alongside, taking its total funding past £28 million, and it is selling something the cynic cannot wave away as decades out: medical radioisotopes, the materials that more than 50 million nuclear-medicine procedures lean on every year to find and treat disease.

The cynic still gets a point, and Astral concedes it. Fusion power, the grid-scale kind that ends up in your wall socket, remains a long way off, and nothing about this round changes that. The move is to stop waiting for it. The company says it already runs multiple reactors at Technology Readiness Level 9, the top of the scale, the rung that means “proven in an operational environment,” and has booked more than £3 million in revenue from research contracts. That is a genuinely unusual sentence to write about a fusion firm, most of which have revenue of exactly zero and a slide deck about the 2040s.

The physics, briefly, because it is the whole moat

Astral’s approach is called multi-state fusion, and the part worth understanding is why it can be small. It uses lattice confinement fusion, a process NASA validated in 2021, in which fusion happens inside the lattice of a solid rather than in a wispy, superheated plasma that has to be held together by enormous magnets. The company says this lets it reach fuel densities some 400 million times higher than those achievable in plasma, which is what allows a useful fusion rate, it cites more than a trillion deuterium-tritium fusions a second, inside a compact, modular reactor rather than a building the size of a stadium. There is a quarter-century of engineering design behind the architecture and a NASA-laureate physicist on staff, which is the sort of detail that separates this from the field’s usual vapour.

That compactness is the entire commercial argument, because of where medical isotopes come from today. Most of the neutron-produced isotopes that hospitals depend on are made at a small handful of ageing fission reactors around the world. A number of them are due to be decommissioned by 2030, and Astral cites an estimate that this would take roughly 64 per cent of global medical-isotope production offline, with little new capacity being built to replace it. The supply chain is thin, international, and fragile in a way that shows up not as an abstract risk but as a delayed cancer scan. It is also lossy: because these isotopes decay constantly, as much as 87.5 per cent of the activity can be lost to radioactive decay in transit before a dose even reaches a radiopharmacy.

Make it next to the hospital, not on another continent

Astral’s pitch is to make the isotopes locally, in compact reactors sited near hospitals, which both shortens the decay clock and removes the dependence on a fragile global chain. The plan is to run several reactors at full capacity by the end of 2026 and bring isotopes to market in early 2027, starting with workhorses like Iodine-131 and Lutetium-177 and moving toward the harder, more valuable targeted-therapy isotopes such as Actinium-225. The company also wants to revive domestic UK isotope production at the former Berkeley Power Station in Gloucestershire, a decommissioned nuclear site, which is the kind of quietly strategic infrastructure nobody notices until it is gone.

This is where it stops being a science story and becomes a sovereignty one. A country that cannot make its own medical isotopes is a country whose cancer care depends on other people’s reactors staying online and other people’s borders staying open, and Europe has spent the past few years learning, repeatedly, how that ends. Astral is small, and the proof is not yet in: a fusion company claiming near-term revenue and a 2027 delivery date invites a raised eyebrow, and well it should. But a fusion company with reactors actually running, contracts actually signed, and a credible physics pedigree is a harder thing to dismiss than the field’s habitual promises. The proof arrives in 2027, when the isotopes either ship or they do not, and the eyebrow comes down or it does not. For once, fusion has set itself a deadline a patient could live to see.