PHYSICAL DYNAMICS&
MATERIAL INTELLIGENCE
Extends our complex-systems methodology into physical substrates—hardware that embodies nonlinear dynamics, multi-scale processing, and resilience under constraints. We build systems that compute using the physics of their environment.
Analog Signal Extraction
Real-time classification of complex acoustic and vibrational signatures using physical reservoir computing, enabling early detection of critical events and targets before noise is digitised or amplified.
Environmental Field Dynamics
Identifies nonlinear shifts in environmental conditions, turbulence patterns, and structural behaviour, providing early warning of external pressure build-ups or internal system fatigue.
Embedded Autonomy Layer
Maps computational intelligence directly into a platform’s material and mechanical envelope, aligning sensor processing with physical constraints to enable long-duration resilience in constrained or denied environments.
MARCO
A compute-in-the-aperture metasurface that performs analog transformation of RF signals before they enter the front-end—providing early awareness of interference, jamming, and environmental conditions at the physical layer.
MARCO uses engineered unit-cell dynamics—nonlinear scattering, fading memory, and wave-speed response—to convert incident RF into rich analog features directly on the antenna surface, enabling adaptive sensing with negligible power cost.
HyperNIC
A stacked analog front-end module that performs pre-ADC inference on RF signals—delivering jam detection, waveform recognition, and beam-quality metrics without altering existing modem architecture.
HyperNIC unifies MARCO’s spatial analog processing with an HTDF temporal fabric, producing compact, low-rate intelligence at the physical layer while shielding downstream ADC and DSP stages from overload and unnecessary data volume.
HMPR
A multi-domain physical reservoir computer that extracts intelligence from underwater acoustic environments before digitization—enabling persistent seabed sensing at sub-milliwatt power.
Our core technology, the Hydro-Magnetic Parametric Reservoir (HMPR), leverages the coupled dynamics of mechanical resonance and ferrofluidic diffusion to perform heavy computational lifting in the analog domain.
HMPR is a tri-domain physical reservoir computing architecture that processes complex signals through the intrinsic dynamics of mechanical motion, ferrofluid viscosity, and magnetic stiffness. This enables analog preprocessing that reduces digital compute burden by more than 10×, providing low-SWaP intelligence for sensing, autonomy, and embedded ISR systems.
