~35% · ~70%
Mid-band / spectral
Mid-band osc advantage ~35%; sweet-band spectral power ~70% vs best EMA on locked characterization packs — surface-dependent.
ZO evidence · Locked · evidence
Product hub · Soft thermostat · Evidence first · PROMOTION discipline
Zero Overshoot (GSRF Practical)
Deterministic soft thermostat for noisy signals — mid-band ring down, peaks toward a declared normal, so the next stage sees a calmer input. Locked packs. Published failure modes. Not a general tracker. Not a controller. Not a universal industrial-PID composite winner under sealed evaluation. Product surface and deep packs: zeroovershoot.com. Method bake-offs: sealed evaluation (controlled evaluation lab; ZO is a candidate).
What it is
A deterministic log-space soft thermostat for noisy signals: mid-band ring damp and peak compression toward a declared normal. Pre-stage signal hygiene — proprietary · evidence-first · not magic AI.
What it is not
Not a general tracker (tracking MAD loses by design). Not a delay-tolerant controller. Not a replacement for joint servos / PLCs. Not “wins industrial actuators” / universal composite champion. Not a book-a-call black box.
Product proof
Product numbers and packs live on zeroovershoot.com — Labs does not invent new metrics. Surface-dependent characterization; red lights stay public. Deep packs on ZO.
~38% · 49/52
Peak compression
Mean peak-deviation advantage vs best EMA; 49 / 52 dataset wins across characterization families — surface-dependent.
ZO evidence · Locked · evidence
298/298
True-cmd continuous wear
Wear 298/298 continuous (ABB / UR / Franka). Open red lights with the wins. Not a controller claim.
ZO evidence · Locked · evidence
Deep links — zeroovershoot.com
Partner funnel — start self-serve.
Sealed evaluation
Request a sealed evaluation by email; what comes back is a locked evidence pack with its limits stated. Not public multi-tenant SaaS.
Customers (after self-serve): zeroovershoot.com/customers · Labs page: ZO Evidence · Partnership
Lab / Workbench scale
Prior-gen locked characterization (GSRF Audit Workbench era): on the order of ~50,000 evaluation artefacts and ~25 GB packs/ledgers under frozen evaluators. Inventory orientation behind public claim surfaces — not a performance SLA; not “50k tests in 30 minutes.” Performance tables on zeroovershoot.com/evidence.
White papers (GitHub)
Public papers repo: github.com/CBoon99/gsrf-papers (evaluation package included; evaluation ≠ production license).
How to engage
- Download the audit pack — Request Audit Pack
- Optional: run gsrf-bench on your CSV — gsrf-bench
- Reply NDA or sample CSV — info@boonmind.io
- Short live/shadow hardware test — protocol on Partnership