burakoktenli / testbed-rover-governance Public

Authority-Governed Assured Autonomy Rover Testbed. System architecture, governance design, and reproducible artifact package. 37-component hardware platform, interactive governance simulator, 350 structured simulation runs across 7 fault-injection scenarios.

rover-testbedauthority-governed-autonomyassured-autonomydempster-shafersatahmaacarasafety-criticaltla-pluscc-by-4.0
4,500+ lines 12 files CC BY 4.0 DOI: 10.5281/zenodo.19143190
main 12 files · v1.0 · Mar 2026
Zenodo Open Simulator Project Page
testbed-zenodo-paper.pdfResearch paper (10 pages) · Governance architecture, DS equations, simulation resultsMar 2026
testbed-simulation.htmlInteractive governance simulator · SATA-HMAA-CARA pipeline, 7 fault scenariosMar 2026
testbed-project-repo.tar.gzProject repository: Python source, TLA+ spec, tests, CI pipelineMar 2026
testbed-BLUEPRINT.pdfFull engineering blueprint (mechanical + electrical)Mar 2026
testbed-SCHEMATIC.svgElectrical schematic diagram (vector), color-coded by node typeMar 2026
testbed-BOM.csv37-component bill of materials with verified sources and costsMar 2026
testbed-ELECTRICAL.json76 electrical connections with interfaces, pins, and voltagesMar 2026
testbed-CONFIG.jsonFull system configuration with component specificationsMar 2026
testbed-GUIDE.mdAssembly guide with step-by-step instructionsMar 2026
testbed-rover.pngPlatform 3D renderMar 2026
testbed-schematic-photo.pngSchematic diagram (raster)Mar 2026
LICENSECC BY 4.0Mar 2026
README.md

Authority-Governed Assured Autonomy Rover Testbed

This repository contains the complete research artifacts for the rover testbed: an authority-governed autonomous ground vehicle designed for contested environments. The system integrates SATA sensor trust fusion, HMAA authority computation, and CARA deterministic recovery into a single pipeline validated through 350 structured simulation runs.

Publication

DOI: 10.5281/zenodo.19143190
Author: Burak Oktenli · Georgetown University, MPS Applied Intelligence
ORCID: 0009-0001-8573-1667
License: CC BY 4.0 · Version: v1.0 · March 2026

Governance Pipeline

  • Stage 1: Sensor Inputs (LiDAR, ToF, IMU, Camera, Wheel Encoders)
  • Stage 2: Sensor Fusion (cross-sensor consistency + disagreement detection)
  • Stage 3: SATA Trust (per-sensor trust + weighted Dempster-Shafer fusion)
  • Stage 4: Mission Planner (path planning under authority constraints)
  • Stage 5: HMAA Authority (trust scalar to authority levels A3-A0)
  • Stage 6: CARA Recovery (GREP phases if authority enters lockout)
  • Stage 7: Command Gate (clamp commands to authority envelope)
  • Stage 8: ESP32 Controller (real-time actuation + watchdog + E-stop)

Hardware Platform

  • Dual-compute: Raspberry Pi 5 8GB (autonomy) + ESP32-DevKitC-32D (safety controller)
  • 37 hardware components, 76 electrical connections
  • Slamtec RPLIDAR A1M8, RPi Camera v3, Adafruit BNO085 IMU, VL53L1X ToF
  • LoRa 900MHz telemetry, ATECC608B secure element, TPL5110 watchdog
  • Total platform cost: ~$484

Simulation Results

  • 350 structured runs across 7 adversarial scenarios (50 per scenario)
  • Zero unsafe actions across all runs
  • Authority transitions with hysteresis compliance
  • TLA+ verified safety properties (48,751 states)
  • Statistical methodology: G*Power analysis, Bonferroni correction

Related Work (Same Research Program)

  • SATA: 10.5281/zenodo.18936251
  • HMAA: 10.5281/zenodo.18861653
  • CARA: 10.5281/zenodo.18917790
  • ADARA: 10.5281/zenodo.19043924
  • MAIVA: 10.5281/zenodo.19015517
  • FLAME: 10.5281/zenodo.19015618

Author

Burak Oktenli
Georgetown University, M.P.S. Applied Intelligence
ORCID: 0009-0001-8573-1667
Website: burakoktenli.com