Conceptual design of a multi-ability reconfigurable unmanned aerial vehicle (UAV) through a synergy of 3D CAD and modular platform planning

Reconfigurable systems are a class of systems that can be transformed into different configurations, generally to perform unique functions or to maintain operational efficiency under distinct conditions. In this paper, we perform the conceptual design of a new offline- reconfigurable Unmanned Aerial Vehicle (UAV) platform that can take up both a quadrotor UAV (QR-UAV) configuration and a fixed-wing UAV (FW-UAV) configuration. Potential uses of such an offline reconfigurable UAV, which provides both VTOL/hovering capabilities and long-endurance/range capabilities, include hazard-critical survey applications in remote locations (e.g., surveying wildfires). This UAV platform is comprised of modules that can be assembled/re-assembled to form the FW and QR configurations (as needed). The entire set of modules are designed to be as light as possible (allowing to be carried in backpacks), and the FW and QR configurations are required to satisfy different endurance specifications (60 mins and 15 mins). A novel combination of exploratory 3D CAD modeling and modular product platform planning is adopted for the conceptual design of the offline reconfigurable FW-QR UAV. Conceived through the 3D CAD modeling process, the QR configuration comprises four ducted rotors, a central pod (with onboard electronics), and a battery; in addition to all the components in the QR configuration, the FW configuration includes flying wing and tail sections, with the ducted rotors mounted under the wing. The modules are attached using a robust lock-pin mechanism, conceived through the 3D CAD modeling process. The modular platform planning process promote the sharing of the maximum number of modules without compromising the performances of the individual configurations, and while minimizing the net weight and net cost of all the modules. The UAV modules are defined in terms of 15 design variables, and the subsequent mixed-integer non-linear optimization is performed using the mixed-discrete Particle Swarm algorithm. Significant design improvements of 41% weight reduction and 38% cost reduction are accomplished through optimization.