Guarded Micro-UAV

An engineering study of a protected micro-UAV that detects release and attempts attitude recovery within propulsion, descent, sensing, and guard-load limits.

Results · Reproduce · Data and figures · Safety

Simulated altitude loss and recoverable tumble rate for three component tiers. The results explain the current gate state; the figure guide records assumptions and generators.

Overview

Midair recovery requires the release classifier, controller, propulsion system, battery, mass properties, guard, and test rig to close as one system. This repository expresses those dependencies as executable models, structured data tables, test contracts, and staged stop/go gates.

flowchart LR
    I[Release cues] --> C[Classifier and state machine]
    C --> R[6-DoF recovery simulation]
    P[Propulsion authority] --> R
    M[Mass, inertia, and CG] --> R
    R --> G[Recovery and descent gates]
    S[Guard and rig checks] --> F[Flight-test readiness]
    G --> F

Loading

The current plots use estimated or catalog-derived parameters. The Monte Carlo study compares placeholder torque with an assumed mixer-authority model. Hardware has not yet supplied the measured authority, mass properties, guard response, or recovery-flight data needed to close the registered gates.

Reproduce the analysis

python -m venv .venv
source .venv/bin/activate
python -m pip install -r requirements.txt
python -m unittest discover -s Analysis/tests -v
python -m Analysis.run_release_recovery
python -m Analysis.monte_carlo_recovery

The first analysis command regenerates the recovery JSON and four plot files. The second regenerates the deterministic Monte Carlo JSON. Other calculators and their inputs are indexed in Analysis/README.md; the complete data path is in docs/data-and-figures.md.

Documentation

Document	Purpose
Analysis/current-results.md	Current mass, guard, recovery, Monte Carlo, and gate results
docs/data-and-figures.md	Data sources, assumptions, plot generators, and reproduction limits
docs/figure-manifest.json	Machine-readable generator/input/output map
Engineering Plan/README.md	Test sequence, dependencies, and exit gates
Design Report/README.md	Architecture, BOM, and preliminary calculations
Controls/README.md	State machine and control interfaces
Instrumentation/README.md	Bench equipment and measurement procedures
Safety/README.md	Release-rig plan and operating controls
docs/specs/measured-authority-gate/	Registered propulsion and recovery acceptance contract
ROADMAP.md	Project phases and remaining milestones

The authoritative controller state names and guards are stored in Controls/state_machine.json.

Repository map

Analysis/          executable models, gates, tests, and result summary
Controls/          recovery state machine and controller interfaces
Data/              generated JSON plus the structure for future test data
Design Report/     BOM, architecture, and calculations
Engineering Data/  requirements, budgets, interfaces, and FMEA tables
Engineering Plan/  staged execution and procurement plan
Instrumentation/   bench procedures and measurement requirements
Safety/            release-rig and operating controls
Figures/           generated plots and authored engineering diagrams
Research/          component, propulsion, and architecture studies
docs/              specifications and figure lineage

Safety boundary

Do not attempt a recovery flight from the current project state. Propulsion measurements, guard testing, release-rig checks, and the registered simulation gates must be completed first. See the propulsion-bench checklist and Safety/README.md.

See CONTRIBUTING.md for analysis checks, generated-artifact rules, and safety constraints. The project uses the MIT License.