Public Lab 03 / Aerial systems
Aerial resilience simulation
A reproducible digital-twin experiment in normalized space: noisy virtual observations, probabilistic state estimation and uncertainty-aware review.
T +0.0/ 6.0 s
Ready. Local deterministic model initialized.
SIMULATED DATA LINK
Synthetic external connection
Orbital telemetry twinLocal VG-SYN/1 loopback · telemetry-in only
- Latency
- 465 ms
- Jitter
- 5 ms
- Packet loss
- 0.9%
- Packets received
- 48
- Data age
- 511 ms
- Integrity
- D60AA084
Boundary: No antenna, orbital service, radio, remote URL, or live ephemeris is contacted. Telemetry is one-way, generated in this browser, and the command channel is permanently disabled.
Academy basis: Scenarios models link conditions; Bayesian Update is represented by the Kalman measurement update; Monte Carlo exposes uncertainty. These are demonstrative public models.01 · NORMALIZED FIELD
Synthetic situation map
The map shows 4 synthetic objects, 4 currently observed, with a normalized RMSE of 0.0885. Synthetic link is nominal. Enabled layers: Truth, Observed, Estimated, Prediction.
03 · PUBLIC AGENTS
Four bounded perspectives
MTH
Mathematics
ActiveValidate deterministic propagation, estimation, uncertainty, and bounded sampling.
FIN
Finance
StandbyInterpret fictitious capacity scarcity without pricing or investment claims.
STR
Strategy
ActiveCompare synthetic priorities, assumptions, trade-offs, and evidence gaps.
MKT
Marketing
StandbyTranslate model outputs into accessible, non-operational public explanations.
This repository exposes no hidden agent network, private prompts or proprietary orchestration.
04 · MATHEMATICS
Estimator health
x̂k|k−1=F x̂k−1|k−1 + wk
Assumptions & validity
Constant-velocity x-y transition over 0.1 s steps; additive zero-mean bounded demo noise; normalized dimensionless space. Cubic Hermite interpolation keeps the render continuous between deterministic steps. The z oscillator and perspective camera are visual mathematics, not measured altitude or a 3D Kalman channel. RK4 and Kalman are domain-specific numerical methods; Academy Scenarios and Monte Carlo provide the public uncertainty context.
05 · MODEL COMPARISON
Same state, different integrator
| Model | 1-step error | Confidence proxy | Cost | State |
|---|---|---|---|---|
| Euler propagation + Kalman | 0.000011 u | 91.6% | 1× relative | Compared |
| Heun propagation + Kalman | 0 u | 91.6% | 2× relative | Compared |
| RK4 propagation + Kalman | 0 u | 91.6% | 4× relative | Selected |
Errors are calculated against the next RK4-generated core state; relative cost describes function evaluations, not wall-clock latency.
06 · MONTE CARLO
Uncertainty distribution
P1044.181
P5045.002
Mean45.115
P9046.194
Distribution varies with scenario noise, capacity stress and seed. It does not estimate real-world outcomes.
07 · AUDIT
Traceable calculation log
- Seeded synthetic observation quality generated.MEAS
measurementCount: 4 · detections: 4 · misses: 0
- Greedy gated association evaluated.ASSO
matches: 4 · unassignedTracks: 0 · unassignedMeasurements: 0
- Constant-velocity Kalman estimates updated.ESTI
updatedTracks: 4 · predictedOnlyTracks: 0
- Visible fictitious risk weights applied.RISK
riskCount: 4 · maximumRisk: 57.171