Studies and Trials

Which technique and mechanics should I use or develop?

Now, I´ve tried to build a mock-up for one side of the handrail.To see which technical problems I have to face.

plane plane plane

Test the movement with Arduino first:

After first trials of a robotic movement I thought about the mechanics how to solve a balancing problem in a very simple way. I wanted to keep the number of joints low, but it was important to stick to a walking handrail. That means only two feet and a stiff pole in between. My former idea to let the robot walk from hole to hole was not anymore the aim I wanted to reach. I would need a lot of fine-tuning, too much for the shortness of time and I just wanted to get my railbot walking freely. The consequence was no footing anymore.

Afterwards I start testing several possibilities for stabilizing the walking balance. My goal was to keep the handrail stiff to be able to use it as an handrail later on. But that means to integrate foot holdings in the construction like vacuum, or magnets or mechanical grasp. As I want to keep it very simple I decided to try first some ideas about geometric solutions, keeping a wheel in the middle of the handrail on the floor.

plane plane

But it was a very unstable equilibrium. So I thought about moving weights from one side to the other.

So, I have tried something in between moving arm up from the middle of the rail and having a wheel beneath. I know it seems ridiculous but I wanted to know if I can find a solution just through geometry without any helping aids in the foot area like vacuum or grasping or magnets. to get a balance back. But it was not successful without changing geometry through moving parts, like weights rolling from one side to another. However, that means more additional parts like joints or pushing elements. That was something I didn´t want as I wanted to keep it simple.


But that means losing simplicity making the very simple idea more complicated.

Then I´ve got the idea to solve the static problems with electromagnets, that would have meant to be restricted to iron metal as surface material.

My next idea was inserting ventilators for pulling and pushing. But it came out that ventilators especially these for computers, are too soft for any strong pushing or pulling power and they run on 12V. After thinking about how to build a strong ventilation I looked around in our lab. More accidentally I saw a package on top of our fablab shelves with brocken copters. I looked into it and saw a big amount of mini quadrocopters. As I need only two rotors I was sure it would be ok to recycle them for my handrail and it came out as an amazing difference even with less voltage (3,7V) then expected. And what is important, too, their poles can be switched that means they can suck and blow.

I started to experiment with geometry, material and these little powerful rotors from Horizon.

plane plane

Motion study for walking/hoovering robot

Below you can see which and how motors and air-screws have to be triggered to get a movement in one direction. Also, the rotation back has to be on the air cussion. If timing is not accurate, the retraction movement of the feet would influence the other. plane