■ SAFE RISK ■ Manufacturing & Production
No. Bell founding is a small, ritualised, high-consequence craft where a single pour that goes wrong costs months, and the surviving foundries are family workshops with knowledge that never got written down.
“Casting bronze temple bells in Japan. Each bell's tone is shaped by fire, metal, and prayer.”
Our AI replacement risk score — how we score jobs
Casting a bonshō is a months-long project with one irreversible moment. The founder builds a strickle-board mould from clay, sand and horsehair, sweeping the profile that determines the bell's wall thickness and therefore its tone. Inscriptions and lotus-band relief are modelled by hand into the mould. The furnace is charged with a copper-tin alloy whose composition the shop guards; the pour has to happen at the right temperature, in the right time, without a hiccup, into a mould buried in a pit. After cooling comes breaking out, grinding, and tuning — shaving metal from the interior while striking and listening, because the long decay and the beat between partials are judged by ear. Then installation at a temple, with a ceremony.
Where automation helps is modelling and machining. Finite-element acoustic simulation can predict a profile's partial frequencies before any metal is melted, CNC can cut patterns and cores accurately, furnace control and alloy spectrometry reduce variability, and industrial foundries do handle large-scale bronze casting for monuments. Those techniques are used by the bigger workshops.
But the market is tiny — a handful of bells a year for temples, memorials and carillons — and buyers are commissioning a named foundry with a lineage, often after a war-metal-requisition history that makes provenance emotionally loaded. Tuning by ear, mould craft in a medium that behaves differently every batch, and ceremonial obligations keep this human. Our score of 8 acknowledges better simulation, not a replaced founder. The failure modes are also unforgiving in a way that rewards experience: a mould that cracks under thermal shock, a cold shut where the bronze froze before filling, gas porosity that ruins the tone. Recognising those risks before the pour is knowledge accumulated across a career, not a parameter set.
Automatability: our editorial assessment of current and near-term AI capability
Resilient, but fragile for reasons unrelated to AI. Simulation and CNC pattern-making will keep spreading through the 2030s and make first-pour success more likely. The genuine threat is succession: these are family foundries with few apprentices, shrinking temple budgets and expensive copper. If the craft disappears it will be because nobody took over the workshop, not because a machine outcast anybody.
Industrial foundries cast large bronze objects routinely, and CNC-machined patterns plus computer-controlled furnaces are used in bell making today. What resists is the last stage: tuning by removing metal from the inside while listening to the partials and decay. That judgement is done by experienced founders, and for temple bells the commissioning party specifically wants a traditional workshop's name on it.
It designs more predictable ones. Finite-element modelling of a bell profile predicts partial frequencies well enough to reduce trial casts, which is valuable when a failed pour costs months and a fortune in bronze. It does not capture everything — mould behaviour, alloy variation and cooling all introduce deviation — so simulation narrows the target and human tuning finishes the job.
It is a shrinking one, though not for automation reasons. Few workshops remain, orders are limited to new temples, replacements and memorials, copper prices are punishing, and apprenticeship is a long unpaid-ish commitment. Survival depends on succession and on diversifying into restoration, monuments and export work. The craft's peril is demographic and economic.
Both halves. Traditional mould building, alloy handling and ear tuning from a working shop — there is no substitute for pours you have watched go wrong. Then the modern layer: acoustic simulation, CNC pattern making, alloy spectrometry and furnace instrumentation, which reduce risk. Add business skills, because winning temple and memorial commissions is relationship work that keeps the workshop alive.