Research project · Clemson University SPRI

Design of Low-Cost Modular Camera Suites for Wildfire Data Collection

Three interchangeable camera concepts designed for use aboard unmanned aerial and terrestrial vehicles during environmental-hazard monitoring and research.

Role: Research intern and designer
Focus: Mechanical design, cameras, electrical design, and embedded systems
Platforms: M500 development drones, LEO rovers, and a custom platform
Tools: CAD, 3D printing, Raspberry Pi, and bench testing

Modular environmental camera systems developed for drone and rover platforms. Picture of information poster and prototypes. — The project explored several ways to provide wide camera coverage without relying on a single expensive commercial sensor package.

The problem

Environmental researchers need broad visual coverage, but existing integrated camera platforms can be expensive and difficult to adapt across different vehicles.

My goal was to develop a lightweight and modular system that could move between drones and ground rovers. Rather than committing to one mechanism immediately, I developed three approaches with different levels of complexity and three different uses: a fixed panoramic module, a conventional gimbal, and an experimental active ball joint.

Module one

Static panoramic dome module

CAD model of the static panoramic camera enclosure — CAD model of the fixed camera enclosure.

The most simple concept used three fixed cameras to provide perfectly aligned fields of view. With no moving components, the design reduced mechanical complexity and created a system that was easy to manufacture, repair, and operate.

This approach traded high quality cameras for immediate coverage in several directions. It also provided a useful baseline for comparing the added value of the moving concepts.

Module two

Two-axis gimbal

CAD model of the two-axis camera gimbal — The gimbal was designed around a compact IMX219 camera.

Physical testing of the camera gimbal — Bench testing the assembled gimbal and camera system.

The gimbal used two 9 gram servos to aim a single, higher quality camera. Compared with the static module, it required more control hardware but reduced the number of cameras needed and allowed an operator to inspect a particular area on demand.

Building and testing this version exposed practical issues that were less visible in CAD, including cable routing, servo range, power delivery, and keeping the drone stable while the mechanism moved.

Module three

ABENICS active ball joint

The most ambitious concept adapted the ABENICS spherical-gear principle into a camera mount. Four servos actuated a compact ball joint, allowing a single camera to point through a much broader 3-axis 360° range than a conventional two-axis mechanism.

This design offered the greatest theoretical coverage, but it also introduced tighter tolerances, more difficult fabrication, and a substantially more complicated control problem. It served as the project’s experimental option rather than the simplest deployable solution.

Results and tradeoffs

The work reduced the estimated cost of wide-area visual coverage from roughly $15,000 for a commercial integrated platform to about $300 across the three prototype approaches.

The static and gimbal systems were bench-tested successfully. The lower cost came with clear compromises, particularly lower image resolution and shorter transmission distance than the commercial reference system.

More importantly, the project showed that one universal camera mount was not necessarily the best answer. A modular interface allowed the vehicle and research team to choose the appropriate balance of coverage, complexity, weight, and cost for each mission.