On 18/12/2015 6:22 PM, Mike Borgelt wrote:
Also a group at Sydney University I briefly communicated with 3 or 4
years ago.................so we put these algorithms in to a
microcomputer, hook it to an autopilot and enjoy the ride?
This is all doable right now. It will be interesting to see if/when the
AI beats a human pilot.
One of my workmates is very active in this area. He's working on
breaking the distance to goal record, which is currently about 300km.
Pretty sophisticated stuff - 4m wingspan glider is the base, mostly
arduino for compute power etc. I've been helping him a lot with various
aspects of the soaring side - even loaned him copies of the original
Riechmann book and Bernard's latest edition. Been talking a lot about
soaring theory and what glider pilots look at in the sky and the ground.
His biggest issue is compute power on the image processing. Lots of
experimentation with thermal processing etc to see if he can detect
temperature gradients in the sky to "see" thermals etc. Quite a
fascinating project all up, and something that will eventually feed into
instrumentation that human pilots can make use of.
--
Justin
_______________________________________________
Aus-soaring mailing list
[email protected]
http://lists.base64.com.au/listinfo/aus-soaring
