Hi, I'm Katie Skinner, I'm a PhD candidate
at the Robotics Institute at the
University of Michigan. So the goal of my
dissertation research is to advance
perception of autonomous underwater
vehicles. Over 70% of the earth is
covered by water, but there's still so
much that we don't know about the sea.
We actually know more about the surface
of Mars than we do about our own
seafloor and I want to change this.
Robots and autonomous systems can have a
key role in telling us more about
underwater environments. You can design
robots that go to the deepest parts of
the sea, and we can develop robots that
can carry out large-scale efficient
surveys for environmental monitoring or
search-and-rescue tasks. The imagery we
collect underwater looks much different
than images that we can gather in air, so
because of water column effects like
attenuation and back scattering, these
images are actually very degraded and
this has a large impact on applying
computer vision algorithms that we
developed for in-air applications to
underwater environments. My research
focuses on restoring underwater images
so that we can improve the way that we
generate large scale 3D reconstructions
or maps of the seafloor, and currently
I'm applying this to coral reef systems
to be able to provide accurate 3D models
of coral reefs to marine biologists who
might want to study these systems, and
ultimately, I want to develop technology
that can enable us to learn more about
subsea environments to gather a
large-scale 3d reconstruction and
high-resolution maps of the seafloor, and
then to be able to automatically
interpret what we're seeing from these
maps to allow us to make informed
decisions about how we should interact
in these environments. I've been able to
work with many wonderful people and
organizations during my time in the
Robotics Institute at Michigan. So my
research is focused in the Deep Robot
Optical Perception Lab, which is advised
by Professor Matthew Johnson-Roberson,
and through my work in the DROP Lab, I've
been able to collaborate on many
interdisciplinary international teams of
researchers who are really experts in
their fields, from marine archaeology to
marine biology. And being able to
collaborate across disciplines has
really transformed the way that I view
my research, and it's changed the way
that I think about problems I want to
solve.
At Michigan, it's so easy to
collaborate and work across different
labs and different departments, and in my
time here, I've been able to collaborate
with the Ford Center for Autonomous
Vehicles, which is focused on advancing
perception for self-driving cars. And
even though this seems very different
than my work in marine robotics, I've
actually found a lot of parallels
between the challenges that we face on
land and the challenges that we face
underwater, and being able to to work
across departments and in different labs
has really enriched my experience here
at Michigan. Robots and intelligent
systems have great potential to push the
bounds of what we're currently capable
of achieving, so I hope to serve society
by developing technology that can
improve safety for tasks that pose
great risk to humans currently, or that
enable us to perform new tasks where
humans can't actually go right now, from
the deepest parts of the oceans to
exploring planets in space. So my
favorite part about my Michigan
experience has been the sense of
community that we have here in the
Robotics Institute at Michigan, and I've
been so lucky to be able to watch the
growth of the Robotics Institute, from my
first year, when we had only eight
students, to now this year, when we have
over a hundred students. And even through
this period of growth, we've been able to
maintain this tight-knit sense of
community that's so important to
enabling collaborations that will allow
us to advance society. And, I feel very
fortunate for the connections that I've
made here, and even though it's been a
very challenging experience, it's been so
rewarding and I'm just extremely proud
to be part of the Robotics Institute at
the University of Michigan.
