NUSwan
NUSwan (pronounced as new-swan) is
a team of autonomous surface vehicle (ASV) platform for monitoring freshwater bodies, developed at the Acoustic Research Lab in Singapore. These robotic swans measure and send various physical and biological parameters of water to a cloud system in real-time. Thus a cooperative team provides both temporal and spatial water sampling capability, while keeping track of water quality parameters such as harmful algal blooms (HABs) that affects the health of a water body's fauna.
Bumblebee AUV
I had a great time working with the Bumblebee autonomous underwater vehicle (AUV) team, during my Masters at the National University of Singapore. Our team was well aware of Transdec pool's murky waters, where the Robosub competition is held annually. This competition attracts many enthusiastic AUV teams from around the world to attempt a set of vision, manipulation and acoustic localization tasks.
We were excited to use a multibeam imaging sonar (Blueview P450) in order to have a better vision capability, supplementing the frontal-camera vision of our vehicle. As an initial step of cleaning the sonar data, a mixture of spatial noise filters like Gaussian low-pass filter, temporal persistence filter, and a normalized power-law filter was applied. These filters came in handy to eliminate the noise, to an extent, caused due to surface and bottom reflections of our testing arena - a 5m deep swimming pool. Objects, placed few tens of metres away from the vehicle, could be detected and tracked pretty well. This cool sensor played a major role in fetching us the runner-up position at the Robosub competition in 2015.
