978-1-4799-5743-9/14/$31.00 ©2014 IEEE Setup of a communication and control systems of a quadrotor type Unmanned Aerial Vehicle Guadalupe Crespo, Guillermo Glez-de-Rivera, Javier Garrido Human Computer Technology Laboratory (HCTLab) Univ. Autónoma de Madrid, Spain guillermo.gdrivera@uam.es Roberto Ponticelli Robomotion S.L. Parque Científico de Madrid Campus de Cantoblanco, C/ Faraday, 7 28049 Madrid, Spain Abstract— This paper outlines the communication and control systems of a quadrotor type Unmanned Aerial Vehicle. The communication system comprises two different links, one for data and other for video signal, and all integrated by an autopilot module. The data link will be implemented using XBee modules and Mavlink communication protocol. The video transmission system will consist of two separate links which can work together avoiding interferences by using orthogonal polarization each other. It is also given some insights of the onboard software architecture, which is based on the Pixhawk autopilot. Keywords— UAV, quadcopter, autopilot, XBee, Mavlink, Pixhawk. I. INTRODUCTION UAV (Unmanned Aerial Vehicle) or Drones are commonly employed in tasks where they have easy access to places that people do not. Due to their small size UAVs can be useful in hazardous conditions where human life is at risk and surroundings that are inaccessible to reach [1]. Its mobility is of key importance in several fields as surveillance, military application, documentation and monitoring purpose and image acquisition [2] [3] [4]. Nowadays, the most common UAV type for non-military application is the multicopter, and particularly the quadcopter [5] [6] which now have sufficient payload and endurance for this kind of tasks [7]. Among the advantages of copters over planes on these tasks is their ability to hovering. The multicopters offer, on the other hand, more payload carrying capacity than a helicopter of the same wingspan. It is known that multicopters controllability is better as there are multiple thrust vectors which renders the flight control simpler, allowing to hover at a constant level from ground by itself and at the same time allowing anyone to easily maneuver it [8], or also to achieve autonomous take-off and landing with little control software development effort from the autopilot point of view. In general, they implement several flight modes, from less to more autonomous operation. The typical flight modes are: - Manual: Roll, pitch, yaw and throttle controls are feed directly from the user input to the autopilot to calculate servo output values in an open-loop manner. This mode should not be available for multirotors. - Stabilized: Roll, pitch, yaw and throttle controls are feed as setpoints to the autopilot, calculating servo output values in a close-loop with the attitude controller. Controls are in manual mode but the aircraft attitude is stabilized in the 3D space. - Autonomous: The aircraft follows GPS waypoints set by base station. It is needed both specialized hardware and software on a flight controller to achieve autonomous flight. In the case of hardware, IMUs (Inertial Measurement Units) are the most relevant module. An IMU is an electronic device that measures and reports on a craft's angular velocity , orientation , and gravitational forces , using a combination of three orthogonal accelerometers , three orthogonal gyroscopes , and three orthogonal magnetometers , enabling to estimate the UAV attitude. Quadcopters nowadays, besides the IMU, are also provided with barometers, vertical range sensors, airspeed sensor and GPS to obtain global position and make possible flying over waypoints. Any variation on position, orientation and acceleration is detected by hardware devices, but data must be processed by the software onboard the craft to realize autonomous flight. There is great variety of projects where flight control algorithms are developed, as for example, those at the University of Zurich [9] and the Flying Machine Arena [10]. They make micro aerial vehicles (MAV) that even are able to do acrobatic flight. When the flight mode is not autonomous, quadcopters need to receive control commands. In most of the commercially available devices, these commands are sent using RC systems such as the widely employed Futaba system [11].