We take hardware to the edge of failure, then write down exactly where it broke.
Rock is the validation arm of Geniotek. We run HALT/HASS, shock and vibration, thermal cycling and ingress testing on production-intent hardware, then hand back the data with the failure mode, the margin and the fix already attached.
Four stress families, one report
Every campaign uses the same scaffold: instrumented units, a bench script with defined stimulus and dwell times, and a report that separates what failed from what merely drifted.
Accelerated life
Step-stress to the first failure, then a production screen derived from the margin you actually have. We stop at the margin, not at the spec sheet.
Transit and field abuse
Sinusoidal sweep, random PSD, and half-sine shock to 40 g. Fixture design is part of the work, not an afterthought handed to the customer.
Cycle and dwell
−40 to +85 °C at programmable ramp rates, with per-unit thermocouple traces so a cold joint and a cold chamber are never confused.
IP69K / IK10
Jet and immersion per IEC 60529, plus impact energy at the enclosure corners where the seal actually gets stressed.
Instrumentation
Every live channel is recorded and exposed read-only during the run, so a failure at 02:00 does not have to be reconstructed from memory.
Failure-first format
Page one is the failure mode list. Margin tables, raw traces and the recommended containment follow. No narrative before the data.
Short campaigns, fixed checkpoints
-
Intake and fixture review
We take the unit as it will ship, review the mounting plan against the intended duty cycle, and flag anything that would make the result unrepresentative before a bench hour is spent.
-
Instrumented build
Thermocouples, accelerometers and rail probes are placed by us. Units are given a run ID and a per-channel log from the first second of stimulus.
-
Run and observe
The bench runs to its script. Live channels are available read-only over the site's telemetry endpoint; nothing in the run depends on someone watching it.
-
Failure triage and report
Within five working days: failure list, margin table, traces, and a containment recommendation for each item, ranked by reproduction rate.
Three ways in
Single campaign
One stress family, one revision. Fixed fee, defined bench hours, report included. Usually the right way to find out what a supplier's first article actually does.
Revision ladder
Same unit across successive revisions until it clears its margin target. We hold the fixture and the baseline traces between rounds so the delta is real.
Production screen design
For units already in the field: derive a HASS profile from the HALT margin, then validate false-positive rate before it goes on a line.