Part of ggSkyGrid · See the connected roadmap →
ggSkybit
A custom quadcopter for exploring airframe design, embedded firmware and flight control.
Objective
Build a custom quadcopter to learn flight control, as a separate branch of the ggSkyGrid autonomy path.
Project Overview
A custom-built quadcopter for learning airframe design, embedded firmware and flight control hands-on. This is an optional branch of ggSkyGrid. The main autonomy path uses a Crazyflie for early physical flight, so it does not depend on this custom controller.
The name is a Tron nod: Bit is Flynn’s little hovering companion. The milestones below describe the custom build if I return to it.
Roadmap & Development Phases
Conceptualization & Architecture
Avionics layout, power distribution budgeting, and PlatformIO toolchain selection.
Airframe CAD & Structural Design
3D-printed and carbon-fiber frame design in SolidWorks/Onshape with optimized thrust-to-weight ratio.
Remote Controller Hardware
Custom handheld transmitter enclosure, dual-axis hall effect gimbals, and battery integration.
Flight Controller Bench Build
Arduino/ESP32 sensor fusion (IMU, barometer) and real-time PID attitude stabilization firmware.
Controller Firmware
Low-latency telemetry communications, channel calibration, and failsafe watchdog logic.
Chassis Assembly & Wiring
ESC wiring, brushless motor mounting, vibration isolation, and physical hardware assembly.
Hover & Flight Verification
Tethered thrust tests, dynamic PID tuning, telemetry logging, and autonomous return-to-home.
Swarm Sim-to-Real Transfer
Flying certified multi-drone plans from the shared planning core, as one unit of a small swarm.
Technical Approach
- Frame: Custom chassis designed in SolidWorks/Onshape
- Flight Controller: Arduino/ESP32-based with PlatformIO toolchain
- Control Software: C++ firmware for motor control, sensor fusion, and PID stabilization
- Remote Controller: Separate custom-built controller with its own hardware and software stack
Connection to GG Swarm
The rover’s firmware, IMU and radio-link lessons carry into this custom flight-control project if I return to it. The main autonomy path instead starts physical flight with one Crazyflie. Swarm work builds on a checked planning core and repeatable single-aircraft missions.
Project Log
1 entry-
2026-03-10 #ggswarm
Week 9 — Isaac Lab Integration and Brain Development
Set up the Isaac Lab development environment locally and got my first Crazyflie drone running in simulation — laying the groundwork for multi-agent training.