The Vehicle Abstraction Layer: A simplified approach to multi-agent, autonomous UAV systems development

Unmanned aerial vehicles (UAVs) are gaining in popularity across a broad spectrum of applications and fields. Two of the main challenges faced when developing autonomous, multi-agent UAV systems is the software development and simulation cycle and the flight test cycle. Each of these aspects can be time consuming and introduce other difficulties, ultimately hindering the mission design and development cycle. In order to address the drawbacks of conventional UAV systems development, the vehicle abstraction layer (VAL) is being developed as part of the Quanser Cooperative Control Framework (QCCF). The goal of the VAL is to provide a robust, portable, multi-agent cooperative control rapid prototyping environment for unmanned aircraft. The VAL allows for vehicle simulation as well as actual flight testing using the same framework, alleviating the need for iterative in-field testing and minimizing the transition from simulation to flight. The VAL enables high-level decision making and control design for multi-vehicle control, high-fidelity simulation and seamless integration on any real Commercial Off-the-Shelf (COTS) UAV system. The framework is open with existing modules (such as vehicle models) available for use as templates for the addition of user-supplied components.