A processing circuit grinds at the pace of the last fix nobody wrote down.

Three in the morning, the SAG trips. The mill coasts down, the cyclones lose feed, and the flotation bank behind it starts running on what's left in the sump. Up the line the crushers keep feeding a stockpile that was never sized to hold a stalled circuit, and the conveyor ahead of the mill stacks ore against a wall it isn't meant to reach. The fitter on shift has seen this trip before, or someone has, but that someone is on days off, or two rosters ago, or at a sister operation a thousand kilometres away. So the circuit stays down while the floor works back from scratch to an answer that already exists.

That hour isn't a repair. It's a search. And on a high-throughput plant, search time is the most expensive time you own.

A comminution circuit is coupled end to end, so the machines fail together. A tripped SAG idles the gyratory and cone crushers feeding it and starves the flotation cells behind it at the same moment, and the cost of a stop is the whole circuit, not one unit. On a high-throughput plant it climbs into the tens of thousands an hour before you count the surge bins backing up, the stockpile you have to claw back through, and the slow ramp to bring a mill back up to load.

The repair is rarely the hard part. The diagnosis is. Most stoppages on a mature circuit are not new faults. They are a cyclone choke, a slurry pump gland gone dry, a bearing alarm someone already chased down on another shift or another circuit, whose answer never made it past their own memory.

Right now the answer lives in the heads of the fitters who have run the circuit longest. It leaves on the next crew change, and it walks off site the day they retire or move to the next operation. A logbook only helps if the next person reads it; a work order tells you the mill went down; what the fitter actually did to clear the SAG bearing alarm or unblock the cyclone is a note nobody writes up in full at 2am, and nobody can find later.

Acervas captures the fix the moment it's made — voice, photo or text, on the circuit, hands still dirty, in seconds. The next time that fault appears on any shift or circuit, the last fix is the first thing the operator sees. A four-hour cold troubleshoot on a Warman pump becomes "check the gland seal water pressure" — the diagnosis you already paid for, handed straight back. Acervas isn't a CMMS and doesn't replace one; it's the knowledge layer on top, pointing each fix back to the work order.

And because the same Metso SAG mills and Weir slurry pumps run at operations you will never visit, a fix logged at one company can answer the same fault at another, anonymised: company, site, circuit and the name of whoever solved it are stripped out before anything travels. You opt in, you both give and get, and your own plants' fixes always rank first.