The idea for creating a hexapod robot was formulated on the field while testing Zippy 2 in Spring 2008. Previously, we had only built robots with wheels and we wanted to improve our mechanical skills by taking on a complex challenge.
Design:
There are three legs on each side of the robot. The four corner legs are mounted at 45 degree angles to the middle legs. Each leg has three servos for movement. The rotator servos are mounted to the base plate of the robot. When they turn, the legs move either forward or backward. There are also hip servos and knee servos on each leg that allow them to extend and retract like human legs.
A lot of thought was put into the leg design. We had a few ideas ranging from one servo controlling all leg movement to three servos per leg. One possible design included two servos with a rotator and hip servo connected to a four bar linkage. We decided to go with the three servo design since it enabled more complex movements. For example, the hexapod is able to use its two front legs to clap. It can also wave a leg in the air to greet us.
The Build: (Detailed Description)
The construction of the hexapod started during the summer robotics camp that was held at Ram's house. While teams were busy assembling and wiring their robots after the day's lesson, we designed and prototyped our first leg. In order to manufacture the parts necessary, we purchased a mini mill. Once a part was milled, a belt sander was used to smooth the edges, fine tune the faces of the plate, and round the corners.
Once the prototype was complete, we mass produced all six legs. This happened at Sergiy's house since Ram was in Las Vegas for the weekend. The day he got back, we moved the mill and belt sander back to Ram's house. That same day, we cut out the baseplate with a jigsaw and mounted all six legs to it.
After that point, it all came down to programming. We worked during the robotics camp at Ram's house most days of the week. Once school started our progress slowed. We had problems giving the legs constant power because as the batteries discharged they would output less power. After trying different batteries and AC/DC converters, we finally ended up buying an expensive power converter that was able to deliver substantial current at 7.5V. After this, programming the robot to walk took no more than a few hours. For a full description of the build and most recent progress updates, check out our Hexapod Blog.
