Our walking motions were made possible by a single motor driving the trajectory of three leg joints via a complex seven-bar linkage system. Two motors in this configuration allows us to “tank drive” the robot to move and steer. The head can tilt or nod toggled by a small solenoid that engages a clutch. Our arms can either come together in a hugging motion or one arm can raise up to wave or hand-hold. These motions were made possible by driving a set of cables from a cam and linkages housed in the chest. I worked on the chest-arm sub-team and was responsible for designing much of the cam and linkage system that drives the arms. By using a cam system, we were able to isolate each arm motion to a half-rotation of the cam by placing the followers accordingly. The motion of the followers is then multiplied with linkages and used to actuate a cable that moves the arms.
This project was only intended to be a design exercise. In order to actually make this robot, we would need to do a lot more testing and review of our designs. I learned a lot about cams, linkages and cable routing in this project but most importantly, learned about tradeoffs between mechanical and control simplicity. Our design would be dead-simple to control because of the mechanical intelligence “programmed” into the mechanisms but for that very reason, it was a challenge to design and build. Since this experience, I’ve been cognizant of finding the ideal balance for each design I make.
In this project, I worked on a team of six seniors in the Mechanical Design course at Olin. At the time, a futurist from Intel was working on a concept called the “21st-Century Robot” or a robot that would be expressive, easily modifiable and manufacturable by the average person. As digital manufacture in the home had been gaining traction, the ability to manufacture your own digital-physical creative platform in your home was a powerful idea. The team at Intel had already designed the look and feel of the base robots and we were tasked with making them move. Jimmy is the first of the proposed series. Additionally, we were constrained to a tight price point which limited us to only four high-torque servos. We needed to be capable of arm and head movements as well as walking. Ultimately, we were capable of attaining two separate head motions, two separate arm motions, forward and reverse walking, and turning - with only four motors.