Radio frequency
Seeing through the barrier
We have developed a unique RF imaging and sensing system for concealed objects, voids and threats. Safely see-through barriers with centimetric resolution and measure structural and material properties of objects.
Our technology
Scyan, our in-house RF imaging system
Scyan is a compact, low power, sub-nanosecond RF sampling system, built entirely in-house with no university involvement, so the IP is fully our own. It runs on microprocessor hardware as simple as an Arduino or an Mbed board rather than the sophisticated high-speed sampling electronics competing approaches need.
In-house scanning technology
- Near field RF imaging, purpose-built for sensor systems
- Time of flight measurement, as in radar, or broadband Fourier decomposition for amplitude and phase
- IP fully owned by Scytronix Systems
Simple microprocessor hardware
- 100 MHz to 6 GHz on hardware like Arduino
- Off-the-shelf low-cost MCUs (as found in the STM32-NUCLEO range)
Material identification and imaging
- Metals, dielectrics and liquids distinguished, not just located
- Ground load bearing capacity, water content and density measurement
- Penetration across a wide range of barrier materials
- Imaging with centimetre resolution using radio waves
Full vector RF measurement
- Linear measurement of amplitude, frequency, phase and polarisation
- No spurious harmonics from conventional frequency mixing
- Sub-nanosecond time resolution, pristine phase measurement
How it compares
Four screening technologies, compared directly.
| Technology | Human safe | Cost | Water penetration* | Contactless** | Weight | Portable | Power |
|---|---|---|---|---|---|---|---|
| Scyan | ✓ | Low | ✓ | ✓ | Light | ✓ | Low |
| X-ray | ✕ | Medium | ✕ | ✓ | Heavy | ✕ | High |
| Ultrasound | ✓ | Low | ✕ | ✕ | Light | ✓ | Low |
| Millimetre wave | ✓ | High | ✕ | ✓ | Medium | ✕ | Medium |
* Water layers under 1 m. ** Stand-off detection, transmission through air before the target.
Working demonstration
Set by the bandwidth and the aperture you can carry
Drive the parameters below and the reconstruction responds the way the physics says it must. Behind the barrier sit three features: a pair of conduits, a void, and a made object with internal structure.
Range resolution
3.8 cm
Cross range resolution
7.5 cm
Two way barrier loss
29 dB
Processing gain
+0.0 dB
Classification
Detected
Void separable.
Resolution and wavelength figures are exact. Barrier attenuation uses an indicative two way model scaled with frequency and material thickness, which stands in for the site survey that would precede any real deployment. Bandwidth and aperture also buy real signal margin, not just resolution: a wider bandwidth compresses to a narrower matched-filter pulse and a longer synthetic aperture integrates more pulses coherently, each adding processing gain against the link budget. A "No return" classification means the barrier loss beats even the maximum gain the sliders allow. The note underneath says by how much, and whether frequency or material is the only way back from there.
What the system does
Safe enough to stand in the beam
Illumination at these power levels sits far below anything that would trouble a person standing in the beam, so the technique works at a checkpoint, a doorway or a wall rather than only inside a shielded bay. A reading also carries material properties, not just position: how a surface reflects and how a volume attenuates lets it separate a filled cavity from an empty one, and metal from a similar shape moulded in plastic.
Typical applications
Structural surveys looking for voids and reinforcement, security screening of baggage and packages, ground work where a buried service needs locating before a digger arrives, and threat detection where a person cannot be asked to stand inside a scanner. Anywhere the question is what sits behind something solid, and cutting into that something isn't an option.
Engagement
From a feasibility question to hardware that answers it.
Most RF work starts with a question about whether a measurement is possible at all. Predictive modelling settles that before money goes into hardware, and the same models then set the specification for the front end, the aperture and the processing chain that follows.