Flight-controller carrier for a tube-launched UAV

A tube-launched airframe imposes the one constraint that cannot be negotiated: everything must fit inside the body diameter, and it must survive being fired out of a tube.

Starting from a board with no source files

The program was built around the Cube Orange+ standard, but the carrier needed to be entirely different — different sensors, different radios, different mechanical envelope. What existed was hardware, not documentation.

So the work started with a structured teardown: layer-by-layer photographic record, part identification, dimensional survey and connectivity traced into a netlist. The result was an annotated schematic that separates measured fact from inference, followed by a clean-room re-implementation of the interface with current, sourceable parts.

Rigid-flex, because there was no alternative

Three rigid sections joined by flex, manufactured at Eurocircuits as a single assembly. In a cylindrical body there is no room for board-to-board connectors and harnesses, and every connector is a shock-and-vibration failure point at launch.

The engineering was in the bend-radius calculation and the stack-up transitions — the flex regions carry signals that still have to hold their impedance across a bend, and the fold geometry has to be assemblable by a human being in the right order.

Getting three radios to coexist

Co-existence was planned in the block diagram rather than discovered at pre-compliance. Physical separation, band planning, and a power architecture that absorbs the cellular transmit transient without disturbing the GNSS reference.

Choosing the payload sensor on evidence, not preference

Selecting the ranging sensor needed evidence rather than a favourite, so the market got surveyed properly: 114 LiDAR and time-of-flight sensors from more than 20 manufacturers, compared on range, field of view, update rate, interface, mass, power, environmental rating and — decisively — availability.

Most of the value in a survey like this is negative: it tells you what cannot work, and why, before anyone has spent money finding out. It also becomes a reusable asset — the same comparison framework carries into the next program with the numbers refreshed.

Design decisions that mattered

Dual-source IMU footprints, because a single-source inertial sensor is a program risk. AEC-Q100 graded parts where the vibration and temperature environment justified it. A power path designed for launch-shock survival rather than desk conditions.