We’re really excited to announce at Caltech
this new initiative called the RoAM Initiative. What RoAM stands for is
Robotic Assisted Mobility The idea of the RoAM Initiative
is to take ideas from robotics, from neural control, and put them together
to help people live better lives to help restore mobility, especially
in the context of bipedal locomotion I’m Aaron Ames, I’m the Bren Professor of Mechanical and Civil Engineering and Control Dynamical Systems here at Caltech, and
we’re standing in the Amber Lab. For a long time our central focus has been
realizing dynamic and efficient walking on bipedal robots. That is: how do we
capture the efficient and elegant way that humans walk dynamically
in a robotic context? What’s interesting is in this process of studying dynamic locomotion from a control theory perspective, from a mathematical perspective,
is the application of this to understanding how to make humans walk better.
That is, how do we take what we know about robotic walking for bipedal
robots, and translate that to help people walk again, walk better, walk more efficiently,
and more importantly, restore locomotion for people? Professor Joel Burdick in mechanical and civil
engineering, he’s discovered with his collaborators that if you stimulate the spinal cord,
that they can actually get the legs to start to fire again. They’ve gotten paraplegics to stand with their own legs by stimulating their spinal cords. The first time you see a patient stand up, you’re hooked.
You just don’t go back. I’m Joel Burdick, the Richard and Dorothy Hayman Prof. of Mechanical Engineering and Bioengineering. More than 20 years ago I started working with Professor Richard Andersen here at Caltech on brain-machine interfaces.
We’ve been developing both implantable and now a non-implantable technology
to stimulate the spinal cord That’s a central piece of our approach –
to sort of revive the damaged circuits below the injury and to
kind of trick them into thinking that they’re talking to the brain.
So, part of RoAM now is to go even one step further, to have the robots coordinate
with the spinal stimulation, not only to improve spinal function but to allow
people to go out and actually use the robots in their daily lives. Really we seem to be at a cusp here. We have all these algorithms for robots that had no human involved whatsoever, So now we have to put a human in the loop and think about how to take those
dynamic algorithms with the human playing a part. What we hope to do is coordinate how the
stimulation changes in real-time with how the exoskeleton is actually moving So as you’re moving through the gate cycle,
when you lift your leg up we’ll probably want to adjust the
electrodes a little bit to give a little more extra oomph to the leg that’s still
standing, and also to encourage the leg which is now in the swing phase to
more naturally go forward And so by hitting them both with the robot side
– which is the mechanical part – and the stimulator – the electrical part –
we can then make sure the circuits are learning the proper
coordination as well. What we’re trying to do is break boundaries. When you start talking about putting a robot on a person, you have to change the paradigm of what you think about what a robot is. And with that comes new science at every
level of the robotic system, and the way it works with the person
to achieve the ultimate goal We can actually make a positive effect on people’s lives
in a day-to-day basis. Their quality of life can improve
if we can restore mobility for them. I think this deserves, and in fact, requires,
our intellectual effort to try to solve this problem.