- Products and Services
- News & Events
Aviating a model helicopter is challenging. A human pilot has to practice hard to be able to stabilize a model helicopter in the air. Currently, there are existing solutions with powerful helicopters however they mostly have heavy and expensive sensors and control-systems on board. The University of Salzburg has developed and built a test platform with an ultra lightweight micro-air-vehicle (MAV) that is controlled by a personal computer. The test platform is implemented in hardware and software with an approach using Ubisense ultra-wideband real-time location system. Ubisense is used to track the position and orientation of the MAV in 3D. A motion sensing device is used for human computer interaction to send flight commands by moving the hands.
DevelopmentThe micro-air-vehicle (MAV) hardware consists of a commercial available cheap four-channel MAV with the capability to control it in each degree of freedom by the software running on Salzburg University PC.
The information about the current position of the MAV in space is measured by Ubisense. The measured position and orientation of the MAV is received over Ethernet.
There is a control system running on the PC which consists of an autopilot algorithm that calculates control commands for the MAV and a reference position. There are two possible modes: in mode one the MAV hovers at dedicated reference positions. In mode two, the calculated control commands are also affected by the motion input of the human pilot. The control algorithm consists of a controller, filter, mode switches and the input of human movement commands. The human commands are received by the PC via Bluetooth from a motion-sensing human-interface-device (HID). The commands by the controller are restricted by virtual walls which prevent the system and environment from damage by irrational human commands.
Results The results show that it is possible to implement an autonomous MAV in a very small and lightweight way. The autonomous hovering enables a robust protection of the surrounding environment by virtual walls. The virtual wall concept increases the possibility to calculate much more complex wall structures in real-time. The system is of a tremendous benefit to the human pilot as the controller balances out flight disturbances such as wind to concentrate on flying directions.
The results showed a real-time delay of approximately 500ms and it was possible up to reach a 10m distance between the sensors to get a consistent signal.