Phase 01
Signal Capture
Get real RF data flowing. SDR spectrum monitoring, WiFi metrics, windowed sampling, and a normalization pipeline for future sensors.
01
Passive RF Environment Awareness
See the invisible field.
Robots.fm is building passive RF sensing infrastructure that reveals structure inside electromagnetic fields already surrounding every space. Signal ingestion, disturbance modeling, and spatial visualization — from desktop prototype to full environmental awareness system.
WiFi disturbance sensingRF spectrum awarenessField-first modeling3D-ready architectureCounter-sensing research
See RF disturbance detection in action.
Interactive preview of the monitoring interface. Watch sensing modes cycle through baseline scanning, active detection, anomaly response, and counter-sensing.
SCANNING
RF Field Monitor
Mode
Zone Status
Living Room
Kitchen
Hallway
Bedroom
Entrance
Exterior
Anomaly0
Signal82%
Events
Waiting...
Real-time visualizationMulti-zone monitoringAnomaly scoringCounter-sensing demo
Core Goal
Investigate ways to become invisible to WiFi-through-wall motion capture.
If ambient RF can be used to infer bodies, movement, posture, or presence through walls, then counter-sensing and obfuscation become a massive defensive research track — not a side quest. The sensing race and the counter-sensing race will grow together.
Signal field concept2D now / 3D later
Primary Mode
WiFi disturbance sensing first, RF spectrum awareness second. Start with what's already everywhere before building exotic sensor rigs.
Core Principle
Raw signal ingestion, field modeling, and visualization stay separated from day one. The renderer never touches raw RF data directly.
Design Thesis
Not merely detecting devices — revealing structure inside an invisible field that is already everywhere. Bigger, more flexible, and more interesting.
01
The system is a field engine.
The architecture assumes from the start that 2D visualization is a temporary projection of a spatial model. The pipeline is built around field data, not UI-specific metrics. Capture, normalize, model, render — each layer fails separately and none depend on the others' internals.
Signal ingestionField modelingDisturbance scoringVoxel-ready rendering
Spatial Model Concept
Voxel field from day one.
Even if the first visual is a 2D projection, the underlying structure thinks in spatial cells, temporal change, and projected disturbance volume. The future version gets smarter; the foundation never needs a transplant.
Pipeline / from invisible signal to visible disturbance
RF Sensors
SDR scans, WiFi adapters, timestamped windows
Normalization
Band grouping, metadata alignment, noise floor
Baseline
Learn normal chaos over time, not pretend stillness
Disturbance
Variance, deltas, anomaly confidence, persistence
Field Map
Voxel-ready spatial, interpolation, projection rules
Renderer
2D now, 3D next, always hungry for better data
Passive Sensing
Detect disturbance without emitting a single signal.
Counter-Sensing
If they can see through walls, learn how to disappear.
02
Infrastructure phases. No fake roadmap.
Each phase produces something real, testable, and motivating — while keeping the architecture aligned with the eventual spatial sensing vision.
Phase 02
Baseline Modeling
Learn what normal chaos looks like. Noise floor estimation, variance tracking, anomaly thresholds, and persistence scoring.
02
Phase 03
Disturbance Field
Translate changing conditions into a spatial field: voxel-ready schema, movement trails, confidence regions, temporal smoothing.
03
Phase 04
Visualization Engine
Make the invisible feel alive. 2D interference fields now, volumetric 3D disturbance volumes next. Replay and debug views built in.
04
Phase 05
Pattern Recognition
Test whether the system can distinguish movement classes. Human vs pet vs aerial vs unknown. Confidence scoring, false positive logging.
05
Phase 06
Counter-Sensing
The big one: learn whether visibility through RF can be weakened, distorted, hidden, or confused. Defensive invisibility research.
06
03
Research directions worth chasing hard.
Some are immediate reading tracks. Some are rabbit holes. The point is to separate what powers the prototype now from what needs deeper study.
Near-term tracks
- General RF disturbance observation
- WiFi signal variance sensing
- Baseline modeling in dense home RF environments
- Ambient field visualization mapping
- Local anomaly scoring on consumer hardware
Deeper study tracks
- WiFi CSI (Channel State Information)
- Multipath reflection modeling
- Passive radar using ambient emitters
- Counter-sensing and obfuscation
- Field fusion across spatial nodes
Build the field engine first.
Not the polished product. Not the grand theory. Build the system that can ingest signals, model disturbance, and make invisible structure legible. That is enough to create momentum, intuition, and a real base for the much crazier versions later.
Robots.fm / passive RF environment awareness / research infrastructure