Perth, Western Australia - 6th to 10th January 2014
Getting a really accurate position from an affordable GPS is surprisingly difficult, especially if the GPS is moving rapidly in a UAV, and is flying in Australia where there is no SBAS (satellite based augmentation system) to provide GPS corrections. Errors on the order of 20m are common which can greatly reduce the utility of GPS for many applications.
The usual solution to this problem is to use a differential GPS ground station, but DGPS systems are very expensive, and thus out of reach of most hobbyists.
In this talk Ben Nizette and Andrew Tridgell will present an attempt to build a free software DGPS system written in Python, using affordable GPS devices and the MAVLink communications protocol to allow for fast real-time improvements in the accuracy of GPS positioning for moving vehicles from a Linux based ground station. The system aims to be practical with low computing resources in the moving vehicle, allowing it to interface with widely used hobbyist level autopilots.
Andrew Tridgell is a long time free software developer, who has recently developed a passion for building UAV autopilots for search & rescue, environmental and related civilian applications.
Ben Nizette is a PhD student at the Australian National University who is working on accurate positioning systems, and is collaborating with Andrew Tridgell on building the DGPS solution.
Ben Nizette is a PhD candidate at the Australian National University developing multi-agent robotic systems. He also lectures in Embedded Systems and has several years experience developing the hardware systems right at the boundary of hardware and software.