There's an old saying that "The proof of the pudding is in the eating", which we've always taken to mean that something like food can look great at first glance, but how does it actually taste when you bite into it? Now that we have our MANOI AT01 robot fully assembled we want to start exploring how it actually performs, and how easy, or difficult, is it to program.
We're still in the early stages of our exploration, but what we've seen so far is surprisingly encouraging. First, we setup the robots 'Home Position'. This is the point of reference posture that, if you spend the time to get it right, allows you to exchange motions and programs with other users, and even with yourself further down the road. Then it was time to start creating some simple programs of our own.
Of course, we plan to load up all the sample motions that Kyosho has begun to publish on the official MANOI websites. We'll do that sometime over the next week or so.
But initially we wanted to try our hand at some original motion creation in hopes that it would help us to develop a gut-level understanding of the robot. It's all to easy to fall into the trap of loading someone else's motions then sitting back and admiring their handiwork. There's nothing wrong with doing it, but it doesn't usually provide any hands-on learning.
What normally works the best for us is to jump into the water with both feet and start trying to paddle around. We fail, and sometimes we fail repeatedly, but we haven't drown - at least not yet. When we find ourselves struggling to grok a particular concept or technique, then we take a peek at what other people have done, or we may even crack open the manual. But, for the most part, we tend to 'learn by doing.' At the same time, we tend to take things slow and try not to get in over our head.
So, our initial attempts at making the robot move started off pretty basic. Move the servos around using the Heart To Heart 3 software while creating a map in our brain - CH01 is the head; CH02 is the left shoulder; and so on. Then we created a simple program to have the robot lift his arms while shaking his head back and forth.
Then, wanting to try something a little more challenging, we created a program with two separate loops. The first loop causes the robot to raise and lower his arms and move his head. We setup a counter so that it would go around the loop 4 times. Then it executes the second loop with a different set of arm motions.
Along the way, we also learned how to use the application software to capture positions. The process is very simple, yet quite powerful and easy to use. We are putting together a brief tutorial since we believe that functionality is a major feature of the RCB-3 controller, especially the way that Kondo has made it so intuitive to use.
Heres a short video record of our initial testing (and learning):
Project M: MANOI Posts (English)
Kyosho MANOI AT01 Official Website (Japanese)