Towards an Open Instrumentaion Platform: Getting the Most From MAVLink, ArduPilot, and BeagleBone

What is a drone? What is an autopilot, and just what is an IMU and a Kalman filter?  This presentation describes an open source hardware and software architecture defined by the Ardupilot firmware, the MAVLink message protocol, several layers of user-space software, and various supported hardware devices and peripherals. It will also cover the current capabilities and components of the core software stacks, as well as extended support for different hardware platforms and sensors, computer vision processing, cameras and image tags, as well as specific science applications and related FOSS projects currently underway. The two highlighted projects both suggest more non-traditional (and less mobile) data acquisition applications using these tools; for more typical UAV applications, airframe options and alternative firmware will also be discussed.

Since both education and the scientific process itself are dependent on freely shared and open ideas, including those expressed as program source code or other engineering artifacts, the advantages of an open instrumentation platform should be obvious. Systems based on open standards and free/open source hardware and software can be constructed at an extremely low cost and offer the benefits of supporting both hands-on learning and real science. The current software has already been developed, tested, and used in many application domains including UAVs and other autonomous vehicles (DARPA Grand Challenge), underwater vehicle research, micro-satellites (cubesats), and fixed instrumentation.