Radiant's SCARP Radio Technology

SCARP Update – Technical Deep Dive

Friends, many of you have been patiently waiting on SCARP, and I owe you a deeper look into what’s been happening. Over the past year we’ve quietly delivered about ten units, and in parallel we’ve been pushing forward on design, sourcing, and production. This hasn’t been without challenges, but those challenges underscore why SCARP matters and why it’s worth doing right.

Core Architecture

At its heart, SCARP is an analog-first, RF-disciplined receiver. Unlike software defined radios (SDRs), which depend on wideband digitization and heavy post-processing, SCARP uses carefully engineered analog selectivity and gain distribution to condition signals before they reach the digital stage.

This delivers three enduring advantages:

1. Signal Integrity: Interference, adjacent channels, and overload are handled in hardware, not left to algorithms.

2. Low Latency: Audio and data flow with minimal delay because there is no dependence on brute-force DSP.

3. AI Compatibility: Pre-conditioned signals are exactly what modern AI models expect. Cleaner inputs lead directly to better transcription accuracy, noise rejection, and robust downstream analytics.

(And yes, in case you’ve wondered about the name: SCARP stands for Software Controlled Analog Receiver Platform. That blend of software flexibility with analog discipline is exactly what sets it apart.)

Improvements to the RF Board

The in-process RF board spin reflects the most comprehensive set of improvements so far:

- Grounding: More via stitching, refined copper pours, and better component selection cut spurious coupling.- Component Quality: Higher-stability semiconductors and passives improve consistency across supply and temperature.- Power Regulation: Cleaner supply rails reduce noise floor and isolate analog/digital sections.- Layout Refinement: Critical trace rerouting lowers parasitics and balances return currents.- Strong Interference Handling: We’ve directly addressed local, powerful interferer problems such as a continuous ATIS on 134.025 MHz overwhelming reception of an intermittent UNICOM on 122.700 MHz when listening too close to a high-watt ATIS transmitter. Countermeasures include transformer-coupled audio output, additional RF cleanup, and new gain-bypass options.

Market-Driven Tuning

I’m looking forward to sharing testing data with AI use-cases. Our current goal is to achieve this within 90 days, dependent on the current board spin.

One thing that has become increasingly clear over the past year: SCARP’s role as an AI-ready signal front end is not accidental.  This direction has been driven directly by market feedback.  Conversations with potential users consistently pointed toward the need for clean, pre-conditioned signals that can be consumed directly by AI models. SCARP has been tuned in response to that feedback.AI readiness isn’t just a buzzword. SCARP conditions signals in hardware — setting levels, rejecting noise, and stabilizing dynamics — so AI models can focus on transcription and analytics instead of wasting cycles on noise cleanup. A clean signal in equals better transcription accuracy, fewer hallucinations, and more robust results downstream. SCARP bridges the analog world of RF with the digital world of AI.

Challenges Along the Way

A project like this doesn’t move forward in isolation, and I want to be transparent about a few realities we’ve encountered:

- Tariffs & Sourcing: Tariffs repeatedly forced us to revisit component sourcing and pricing, creating delays that were outside our control.- Radiant’s Growth: Radiant has grown to 2.5× last year’s revenue. That’s good news — it means we’re a strong, sustainable company and the doors stay open. But it has also stretched my personal bandwidth. SCARP remains core to Radiant’s roadmap as I keep Radiant’s day-to-day operations running.- Patience Required: These realities have slowed the pace, but they have not reduced our determination.

Our next focus is validation of the improved RF board, with lab and field testing to confirm performance under real-world conditions. That will set the stage for scaling production.

Why Not SDR?

Many ask, “Why not just use an SDR?” It’s a fair question, and the short answer is: SDRs are general-purpose tools. SCARP is a purpose-built instrument. The difference matters.- Noise & Overload: SDRs often collapse when strong adjacent signals are present. If you’ve ever tuned an SDR near a powerful transmitter, you’ve seen front-end overload, intermodulation products, and false signals appear. SCARP’s analog selectivity and gain distribution are designed to prevent that.

- Form Factor & Dependence: SDRs depend on a host PC or tablet for processing and filtering. That adds bulk, cables, operating systems, drivers, and constant updates. SCARP is self-contained and disciplined — the filtering and rejection happen in hardware, before audio is ever digitized.

- Audio Quality: SDR audio is often “mushy” in crowded RF environments because the digital backend is struggling to correct problems that should have been fixed in analog. SCARP audio is sharp, natural, and intelligible. That difference is not cosmetic; it’s what makes SCARP’s output suitable for AI transcription.

- Power & Efficiency: SDRs digitize huge swaths of spectrum, which burns power and requires more compute resources. SCARP only processes what matters. That makes it efficient, quieter electrically, and more predictable.

- Reliability: SDRs are excellent tools for experimentation and broad coverage, but they are not engineered for the kind of consistent, disciplined reception needed in real-world aviation or commercial environments. SCARP is engineered for that discipline — repeatable, predictable, reliable reception under harsh conditions.For disciplined, reliable reception — and especially for applications where intelligibility and AI readiness matter — SCARP is the right tool, not SDR.

Tentpole Technology

Through all of this, one thing has become clear: SCARP is a tentpole technology for Radiant. It bridges analog RF discipline with the demands of AI applications, opening market opportunities in aviation listening, training, commercial monitoring, and beyond.

Looking Forward

I’d love to hear from you — how do you see yourself using SCARP once it’s in your hands? Your feedback helps guide where we tune and prioritize next.

I know many of you have been waiting well over a year. I respect that patience deeply. SCARP is not a “ship quick, fix later” gadget. Instead, it’s a disciplined engineering platform built to last. The upcoming RF board represents the most comprehensive set of improvements so far.The result will be worth it.Thank you for continuing on this journey with me.— James