A 25-litre pack has roughly fourteen hardware pieces on it that can fail. Side-release buckles on the hip belt and sternum strap. Ladder-locks on the compression straps. Cord-locks on the drawcord. Sliders on the side-access zips and the main clamshell zip. Tri-glides where webbing terminates. The number is not exact. Different load-out configurations push it up or down. But on the wrong day in the wrong conditions, every one of those pieces is a single point of failure between the user and a working pack.
The lab's repair-forward principle, applied to soft goods, comes down to a question that's actually quite narrow: when one of those pieces fails, what does the user have to do to get the pack working again?
There are roughly three answers a manufacturer can give.
The first is the proprietary answer: send the pack back, here's a returns label, four to six weeks. This is the prevailing model. A pack that fails on day three of a deployment doesn't get back to a returns depot. It gets carried, or it gets binned, or it gets jury-rigged and the user makes do.
The second is the branded-spares answer: we sell replacement parts directly. Better, but still wrong-shaped. It ties repair timeline to the manufacturer's stock levels and continued existence.
The third is the open-spec answer: every hardware piece is sourced from a public industry standard family, available from any outdoor or industrial supplier. A failed buckle is replaced with the same part bought from a chandlery or army surplus shop. Repair time is a few minutes with a multi-tool. Repair cost is typically under three pounds.
That third answer is the one Keystone is being designed to.
In practice this means committing to several constraints that pre-determine a lot of design decisions.
Hardware family. Keystone's hardware will come from a single public-standard family, most likely ITW Nexus or Duraflex. Both are stocked across the UK and EU through outdoor and military supply channels. Both publish full part catalogues with sizing and breaking-strength data. The lab is currently evaluating between them and parking the commitment until REV A04 field trial. Decision criteria: (1) availability through UK channels without import dependency, (2) abrasion behaviour over a 1000-hour wear cycle, (3) failure mode. Buckles that crack progressively are preferable to buckles that crack catastrophically, even if they test stronger on the bench.
Webbing. Primary webbing is 25mm to UK MIL-spec. The 25mm size because it's the dominant pattern for this size class, meaning hardware availability is widest and MOLLE/PALS compatibility is preserved. MIL-spec over commercial-grade because commercial webbing is rated under standard-temperature testing only. A pack designed for the operating envelope Keystone is designed for needs to perform at the bottom of the envelope, not the middle of it.
Zips. Main and side-access closures use large-gauge standard YKK zips with full storm flaps rather than Aquaguard or similar water-resistant zips. Aquaguard zips fail under abrasion in ways that are difficult to repair in the field. Standard zips with storm flaps fail more progressively, can be maintained with zip lubricant and occasional slider replacement, and the storm flap is a sewn component rather than a bonded one. A failed slider on a standard YKK is a five-minute repair with a multi-tool. A failed Aquaguard zip is a return-to-manufacturer event.
What the lab is giving up to make the claim hold is also worth naming. Keystone will not be the lightest pack in its size class. Open-spec hardware is heavier than custom-moulded alternatives. Keystone will not have the most weather-resistant closure system. Storm flaps are not the same as bonded zips. These are real costs, and the lab is choosing to pay them. The resulting pack does the thing the lab cares most about: it survives the day after the bad day.
A spares specification list will be published with REV A04. Until then, this note is the lab's commitment to the principle, with the specifics still under evaluation.
RELATED PROGRAMME
Keystone 25 →