What's this course about? Mechanical models and empirical studies that shed light on the coordination, speed, energetics, and stability of locomotion in legged animals (especially people) and robots.
What's this course not about? On the bio side: physiology, chemistry, neurons and vision. On the robot side: path planning, 3D kinematics, vision, neural-nets, fuzzy logic, algorithms, Kalman filters and Bayesian estimation.
What are the big questions? Why is this even a subject? What limits an animal's running speed? Why aren't there any good legged robots? How can you think about the control of complex mechanical systems? Is there a relation between how animals work and how robots (and machines generally) should work?
I don't plan to make a career out of legged things, why should I take this course? To build general understanding of the mechanics of machines, animals and locomotion. To build general quantitative and computer skills. Curiosity.
Is this a robotics course? Yes. This course covers one aspect of robotics, the mechanics and coordination of legs. Some of the ideas may generalize to other robotic motion tasks. For example, juggling is similar to gait in many regards. More generally various topics in the mechanics of robot motion will be covered.
How is this different from other robot courses? A lot of robotics concerns geometry and planning related to geometry. This course needs some geometry, of course, but the emphasis here is on mechanics and dynamics (both F=ma and dynamical systems).
Is this a biomechanics course? Yes. Animals are governed by the same laws of physics as machines. The same ideas are used to understand animal locomotion as are used to design robots. We will talk about `motor control', as in 'large motor skills', discussing issues about how muscles are coordinated.
How is this course different from other biomechanics courses? This is dry biomechanics. No dissections here. This is about understanding gross motions of animals, not about the materials the animals are made of (OK, there will be a bit about the properties of muscles and a teany bit about respiration).
Can a biologist take this course? Yes. The course is aimed to serve biologists interested in locomotion so long as they are comfortable with math and computers.
What do I need to know at the start? A bright senior
or graduate student who is adept at quantitative reasoning and learning computers
can catch up on any topic below that they don't have firmly in hand. A
good starting point would be:
- Any post-freshman mildly advanced dynamics class such as
TAM 203, TAM 570, MAE 571, MAE 479, or AEP 333
- Basic Matlab skills (ODE simulation, linear equation solution, basic graphics commands).
- Comfort with elementary Linear Algebra and Ordinary Differential Equations (at the level of Math 2930 or 2940).
Can an undergraduate take this course? Yes. See above. The course should really have a 5000 course number.
What kind of books cover material like this? Here
are some that have some overlap with this course content.
Principles of Animal Locomotion, R. McNeill Alexander 2003
Muscles, Reflexes and Locomotion, Thomas McMahon, 1984
Legged Robots that Balance, Marc Raibert, 1986
At the end of the course what buzz words will I be able to use? External work, internal work, limit cycle, Poincare map, `modulus of the eigenvalue is less than one', passive dynamics, ZMP, COT, VO2, CPG, eccentric contraction, simultaneous collision, Koenig's theorems, similarity, optimal control, jump condition, Hill's equation, basin of attraction, linear control. You won't be an expert in all of these topics, at least not from what will be taught in this course, but you will know enough to use the words in a sentence and sound coherent.
What will happen in lectures? General introductions. Videos. Detailed explanations and calculations concerning mechanics, energetics, stability, etc. Some programming tips.
How will I spend my time when doing homework? You will read papers and do short mechanics calculations, by hand and on the computer, related to legged locomotion. A main part of the course is developing the capstone project of writing a simulation for a simple locomotion model and being able to analyze its effectiveness, stability and energetics.
What's the work load? For a typical student a three
credit class should take about 9-10 hours/week of your time on average. Three
of these are in class. You should budget at least 6 hours/wk of homework
and study time.
How will I be graded? There will be weekly homework (some developing into projects and a final project), 1 prelim and a final exam. No labs.
Send email with questions or suggestions.