Bird-flight is a very complicated, not yet fully understood mechanism. Any control system hoping to mimic anything resembling the flapping flight of actual birds, will have to deal with many unsteady and non-linear phenomena. Having proper data available regarding its surroundings and e.g. the interaction with the air seems to be a crucial part in controlling flapping flight. Steps have been taken to develop appropriate sensors and getting these sensors into a robotic bird is one of the next steps being worked on.
The plan is to use both traditional flight-data from instruments such as IMU(s) and GPS, as well as utilise many strain gauge based sensors for e.g. wing deformation and aerodynamic measurements. At first only a few strain gauges will be used, but it is expected that this will expand to 20+ in the future.
However, this reveals an issue on the more practical side of building robots with many sensors. The reading, aggregating and processing of the data of many (distributed) sensors, while staying within the contraints of e.g. weight, volume and time-accuracy as well as being scalable, is a challenge in and of itself. Therefore this individual assignment focuses on designing, building and testing a scalable system that could be used to read many strain gauges, while keeping e.g. weight and volume usage low enough to still enable flight.