IAM3D competition drone.

Design around the task the aircraft must complete.

The competition mission required an FPV racing drone to carry a small payload, navigate an obstacle course quickly, align over a target, and release that payload without manual intervention. Speed mattered, but so did repeatable pickup and release. Every gram added to the mechanism reduced flight performance, while every unverified shortcut threatened mission completion.

The team split the problem into four interacting systems: flight hardware, a payload-retention mechanism, a camera-alignment mechanism, and a battery-restraint strategy. Additive manufacturing made custom geometry feasible, but it also introduced print orientation, tolerance, support material, and layer-strength constraints from the beginning.

Give the engineering work a structure it can move through.

As Project Manager, Muhammad directed the four-part plan from early research through design, manufacturing, and competition preparation. He organized and led general-body meetings, introduced project-tracking and design-documentation practices, coordinated procurement, managed the budget for material and tools, and handled the travel planning required to bring the finished project to competition.

That role connected rather than replaced the team’s technical owners. Decision matrices created reviewable component selections. Shared CAD and report standards kept mechanism interfaces from drifting apart. Procurement timing exposed choices that were theoretically strong but unavailable, over budget, or incompatible with the remaining build schedule.

Balance performance at the system level.

The team used weighted decision matrices instead of selecting the individually fastest or cheapest part. Processor performance, voltage compatibility, sensor quality, current capacity, size, mass, radio reliability, video performance, discharge rate, and cost were considered across the complete flight stack.

The selected configuration paired a SpeedyBee F405 V4 BLS 55A flight-controller and ESC stack with a Tattu R-Line 1550 mAh 4S1P 14.8-volt 130C battery, four MEPS SZ2207 2750KV motors, DJI O3 video hardware, and a TBS Crossfire radio system. The design report’s complete bill of materials totalled $496.56.

Choose a mechanism whose safe state matches the mission.

An energize-to-hold electromagnet was attractive because it simplified release, but it consumed power continuously and generated heat while the payload was attached. An energize-to-release configuration removed that continuous demand but introduced custom-circuit and rules uncertainty. The selected direction used a passive permanent magnet for retention and mechanical separation for release.

The magnetic circuit still had to be sized. With an estimated 22.23-gram payload and a deliberately conservative acceleration case, the report calculated a minimum retention requirement of 1.4371 N. The selected magnet was rated at 5.829 N, providing margin before the design moved into its release geometry.

Convert half a servo turn into a predictable separation.

Magnet-force research indicated that the steel target needed to move at least 0.1875 inches from the magnet before attraction fell below the payload’s weight. Because the selected micro servo could rotate only 180 degrees, the lead screw had to produce that travel in half a revolution. The resulting custom pitch was 0.375 inches per turn.

A lead-screw force model then placed the required drive torque at 0.02 N·m. The smallest 2.1-gram servo fell just short, so the team moved to a 4.3-gram servo rated at 0.037 N·m. The final cage kept the magnet fixed while a servo-driven platen created the release gap.

Expand the pilot’s view without adding a second video system.

A second camera and ground station would have given dedicated downward vision, but available goggles could not accept two live inputs and the added monitor, transmitter, mass, cost, and control complexity worked against the mission. The team instead designed a mechanical mount that could reorient the single DJI camera.

The first geared concept could rotate the camera, but it consumed volume, required a two-piece mount, and needed at least half an inch of rear wire clearance. A top-mounted direct-drive concept lowered the camera enough to distort the pilot’s perception of the drone envelope. The chosen bottom-pivot arm kept the normal forward position above the frame, used less material, and swung the same camera toward the payload.

Moment-of-inertia and gravitational torque estimates placed total demand at approximately 0.00186 N·m, low enough for the lighter 2.1-gram servo. Here the analysis supported a simpler actuator than the payload mechanism because the camera load case was fundamentally smaller.

Use geometry to resist the direction of motion.

The battery system compared two straps against one strap with grip tape. The lighter single-strap approach was retained, but friction alone was not asked to control fore-aft movement. Grooves located the strap and raised edge walls created a positive geometric stop against acceleration.

Remove material only after defining the load case.

The selected 55-amp ESC limited the planned motor operating point to 40% throttle. MEPS thrust data translated that condition into a 5.072 N load at each motor. A baseline SolidWorks study produced approximately 0.34 mm of maximum arm displacement, and the team set 0.5 mm as the allowable limit for candidate frames.

Altair Inspire topology studies explored 20%, 30%, 40%, and 50% mass targets. The raw outputs were treated as design guidance—not printable final geometry. Load paths that persisted across the studies informed a manually reconstructed frame with continuous members, printable transitions, fastener lands, and space for the payload and camera mechanisms.

The refined design projected a 27% mass reduction while staying below the displacement limit under the report’s combined loading. That result created a useful trade: less inert mass for the same propulsion system without accepting an unbounded loss of stiffness.

Check the operating band, not just a static snapshot.

Static stiffness could not show whether the motor operating range would excite a structural mode. The team extracted thirty modes and focused the harmonic-response study on the 1152–1779.1 Hz range associated with sustained operating throttle. A response near 1220 Hz was the most significant point in that interval, but the predicted displacement remained below 0.5 mm.

The result did not claim that simulation replaced flight testing. It narrowed the structural risk and showed that the optimized geometry retained acceptable predicted motion across the analysed operating band.

Design the printed part around layers, clearances, and supports.

CarbonX carbon-fiber-reinforced PLA was selected for the structure, and the available Bambu Lab X1 Carbon could process the material. The frame was kept largely planar so it could print flat without support material, increasing bed contact and aligning layers with the primary in-plane loads.

Mechanisms required a different kind of validation. A printed clearance coupon tested multiple offsets and identified 0.15 mm—0.006 inches—as the report’s usable free-moving fit for that setup. Screw holes and moving interfaces were then sized from a measured process capability rather than a generic tolerance.

The magnet cage was printed on its side so layer orientation better resisted the expected shear. Camera-mount orientation balanced support material against the load direction and surface requirements. Slicer previews provided a final manufacturing check before committing time and material.

Separate what the report proved from what competition proved later.

The report documents component checks, electronics setup, mechanism development, structural analysis, printed-fit testing, and final assembly. At the time it was written, the team’s complete flight-test program had not concluded. The page preserves that boundary rather than presenting simulation and bench work as finished flight validation.

After the report, the project advanced through final integration and competition. The team finished fourth internationally. That later result belongs at the end of the engineering story: it reflects a design that moved beyond CAD and documentation into a complete competition program.