# Horizon Flight Computer > Agent-facing counterpart to the [human project page](/projects/horizon-flight-computer/). ## Record metadata - Record: 005 - Slug: horizon-flight-computer - Domain: Hardware - Domain code: HW - Type: Physical build - Status: Archived - Period: 2022–25 - Portfolio role: Flagship - Publication state: Engineering report; case study in preparation - Case-study readiness: Draftable - Compendium edition: 0.6 ## Summary A model-rocketry program spanning custom airframes, launch hardware, sensors, simulation, and flight data. ## Overview Horizon grew out of the decision to learn model rocketry by building more of the system than a complete kit would require. Airframes, printed parts, launch hardware, simulations, sensors, and data analysis became one long engineering thread rather than separate assignments. A high-school engineering thread combining a modular model rocket, custom launch pad, 3D-printed structures, fin and nose-cone experiments, OpenRocket simulations, sensor trials, MicroPython logging, and recorded flight data. Purpose: The project aims to learn aerospace engineering by building and measuring more of the system instead of relying on a complete kit. ## The problem behind the project The project deliberately chose the harder custom path because it created more room to learn about aerodynamics, structures, electronics, programming, calibration, and experimental error. That choice created useful friction. Aerodynamic intuitions had to survive simulation; sensors had to survive vibration and calibration; and attractive graphs had to be checked against uncertain logs and physical events. Failure was not incidental to the project—it was one of its main sources of information. The immediate beneficiaries were Josiah and his teammate as student engineers. A public case study can help other students see the value of documenting failed sensors and contradictory results. ## How it took shape Kraft airframe components, M3/M4 modular fasteners, 3D-printed parts, a custom launch pad, Pico-class electronics, MPU6050 and barometric sensor experiments, MicroPython, OpenRocket, and spreadsheet analysis. The work combined OpenRocket modeling, modular mechanical design, MicroPython logging, Pico-class electronics, inertial and barometric sensors, spreadsheet analysis, and repeated physical iteration. Some paths worked, some instruments produced ambiguous records, and the BME280 route failed rather than disappearing from the history. Josiah’s exact contribution versus his teammate still needs to be documented before the full public case study. Existing records show substantial work across research, design, electronics, programming, simulation, and analysis. The published engineering report documents the airframe, hardware, electronics, simulations, material comparisons, and fourteen technical diagrams. The wider archive also includes build narratives, bills of materials, sensor code, flight spreadsheets, noisy logs, and altitude traces. ## What the project means now The strongest version of the Horizon story is not a claim of polished aerospace hardware. It is a record of learning to negotiate between models, materials, electronics, and evidence, including the discipline to preserve contradictory results and unresolved provenance. Some logs have uncertain provenance, the BME280 path failed, and several conclusions need real-world wind-tunnel or repeated flight testing. Image and teammate permissions still need review. Simulation results can overturn an intuitive hypothesis, and failed instrumentation is still useful when its uncertainty is recorded honestly. ## Future direction The case study will become complete once teammate contributions are confirmed, logs are matched to physical tests, and approved photos and video are assembled. ## Published documents ### The Many Variations on Model Rockets which Result in Greater Altitude A 22-page engineering report covering the Horizon airframe, modular hardware, flight-computer experiments, simulations, materials, and fourteen technical diagrams. - [Read or download the PDF](/documents/the-many-variations-on-model-rockets-which-result-in-greater-altitude.pdf) - [Download the Word document](/documents/the-many-variations-on-model-rockets-which-result-in-greater-altitude.docx) ## Publication and interpretation notes - Current classification: Archived - Portfolio readiness: Draftable - Publication boundary: Engineering report; case study in preparation ## Additional agent context The value of this record includes contradictory simulation results, failed instrumentation, and learning across physical and computational systems. Teammate attribution and log provenance remain open before full publication. ## Related project records - [Universe Screen](/projects/universe-screen/llm/) — A spatial interface for seeing active projects, files, and AI agents as a living solar system. ## Navigation - [Complete project index](/projects/llm/) - [Public Lab](/lab/llm/) - [Human version of this record](/projects/horizon-flight-computer/) - [About Josiah's working method](/about/llm/) - [Agent discovery map](/llms.txt)