Year 2016 has past and we are now in 2017, nearing SAUVC 2017.
There are still many challenges up ahead that we must face. One main issue that we are tackling is computer vision. The colour of the object of interest, which is a red pail (as seen below), does not stay the same with increasing distance underwater.
With increasing distance, the red pail becomes darker. This poses a problem for the software team as they are trying different methods to filter out the pail from a faraway distance for maximum consistency. Currently, we are unable to detect the pail from a distance very consistently. There are many occasions where Hornet 2.0 will fail to detect the red pail. Works are currently being done to improve this.
Another issue we are facing lies in the acoustics section. The hydrophones that we are using limits the distance as to how far we are able to detect the pinger. Huge amount of noise is also introduced when the thrusters are active. The acoustics team are trying their best to filter out the noise to get a better signal from the pinger. They are also working on obtaining the time difference of arrival from the signal that is produced by the pinger. This will help to locate the pinger’s location so that Hornet 2.0 is able to bump the pinging flair and knock the golf ball off the flair.