Abstract: Australia is a large, sparsely populated, resource rich country a long way
from anywhere; and is consequently the ideal place to do field robotics. The
past decade has seen substantial technical development and investment in large
scale field robotics, especially in civilian applications such as cargo
handling, mining, agriculture and marine environments; applications which are
of central importance to the Australian economy. This talk will describe a
number of technical advances in the areas of navigation, perception, machine
learning, large platform control, and systems engineering that have enabled substantial
progress in the “science” of field robotics and which have led to significant
commercial applications. The talk will
also aim to look forward to the next decade, especially focusing on the
development of machine learning methods for real-time operation of robots in
large-scale unstructured field environments and where the opportunities for
future commercial developments will come from.