ScanIQ · Engineering|AIWarehouse · AI media & digital goods ↗

Pilot + Roadmap

Research

Three research programs connect ScanIQ's core geometry/physics engine to resonance optimization, GPU vision and quantum magnetometry. Maturity labels are part of the product, not a disclaimer hidden in the footer.

Pilot

Parametric resonance mapping

Instead of asking for one mode shape, ScanIQ sweeps geometry, material and boundary conditions around the modal solver to build a map: frequency versus design variable, with sensitivity ranking. The practical question is: which dimension moves which mode, and how far?

Rocker linkage resonance mode
Rocker linkage resonance: mode 1 at 4,896 Hz
Shock body resonance mode
Shock body (trestle v5) resonance: mode 1 at 1,121 Hz

Design question

“Tune the part away from the engine's excitation band.” The map makes sensitivity visible before the geometry is committed.

Pilot

GPU photo / video pipeline with NVIDIA DALI

Heavy image streams should not queue behind CPU decode. DALI moves decode and augmentation onto rented NVIDIA GPU pods so reconstruction and AI models can consume frames without the CPU becoming the choke point.

Today

The CPU server handles capture, file sorting, geometry and orchestration. Heavy jobs burst to rented NVIDIA GPU pods billed by usage.

Pilot target

A 300-frame phone video can be decoded and prepared on the GPU so scan reconstruction can move toward minute-scale turnaround on an on-demand pod.

Roadmap · Research

Quantum magnetometry + AI discriminator

ScanIQ's contribution is the software layer: magnetic forward modelling, discriminator training, source/provenance handling and verification. Candidate sensors include optically pumped magnetometers and nitrogen-vacancy diamond sensors.

RoadmapResearch

Track A · Seafloor magnetic fingerprint navigation

A submersible or AUV estimates where it is by matching measured crustal magnetic anomalies to a map, without GPS or acoustic beacons. Current ScanIQ work is simulation and public-data software research.

Forward models · public MagNav data/tooling · AI signal discriminator · provenance
RoadmapResearch · not a medical device

Track B · Fetal heart biomagnetic mapping

Research into separating fetal cardiac magnetic signals from maternal and environmental components, then mapping candidate sources in 3-D. No diagnostic or clinical claims.

OPM/NV research · MNE signal processing · source mapping · AI discrimination
No magnetometer hardware is owned today. This entire magnetometry track is simulation and public-data work until a sensor is acquired. Biomagnetic work requires research partners and remains research only.
Maturity table

What exists today / what's next

No invisible maturity inflation.

ProgramStatusExists todayNext
Parametric resonance mappingPilotModal solver and real mode results.Automated design-variable sweeps + sensitivity ranking.
GPU vision with NVIDIA DALIPilotCPU capture/orchestration; rented GPU burst compute.DALI decode/augmentation feeding reconstruction/model stages.
MagNav AI discriminatorRoadmapSoftware simulation and public data.Acquire/partner for sensor hardware; validate discriminator with real sensor runs.
Fetal biomagnetic mappingRoadmapResearch architecture only.Research partnership, appropriate instrumentation and governed study design.
Helical/toroidal propulsionPilot7-blade geometry + toroidal coil field map.Queued CFD validation and hardware iteration.