Problem
Our assignment is to produce a novel and interesting mechanism. It can be almost anything but we need to demonstrate that we are able to take the tools we have learned in this class and use them to solve a realistic problem. We need to take our analytical solutions and use them to inform decisions that solve larger problems outside of the textbook.
Background
Bipedal locomotion is an exciting area of robotics research. Many different solutions have been developed for the problem of how to make a two-legged robot walk reliably. Most of these solutions are extremely complex, utilizing high-degree-of-freedom bodies and sophisticated electronics and software for motion planning and joint control. The most difficult problem in bipedal locomotion tends to be balancing during motion. Powering the robot is easy, making the robot move through the commanded trajectories is easy, but making it balance while walking is very difficult. We wanted to see if we could find a solution to the problem of producing balanced bipedal locomotion using the materials covered in this class.
Solution
We propose to create a walking bipedal robot using the information covered in this class. It is fairly simple to produce a simulacrum of "walking motion", but much more difficult to actually make a robot walk and carry its own weight. More importantly, it will be difficult to make that robot balance when it has to shift its center of gravity from one foot to the other during each step. Inexpensive toys can demonstrate bipedal locomotion, but only for the special case that the feet overlap and the body's center of gravity is motionless. This is not an accurate representation of the locomotion of any living animal or human-sized bipedal robot.
So our robot must overcome three main obstacles: it must walk to move itself, it must be able to carry its own weight, and it must use its own weight to maintain balance while walking. We cannot imagine a single mechanism that accomplishes all of these goals, but a combination of several complementary mechanisms should suffice.