Humanoid robot A1
ABOUT THIS PROJECT
Tell us briefly about your project. Is it a work of art or does it serve another
The A1 is an adult size humanoid robot. Its torso has two arms, each with five (six including the hand) axes of rotation based on igus robolink robotic joints. Currently, only the right arm is motorized. The arm's joints are tendon driven by stepper motors with planetary gearheads attached. An experimental version of a motor controller is completed. It can read out the rotary motion encoders in the joints, drive a stepper motor and perform position control. The A1 humanoid has a sensor head with two degrees of freedom. The head is equipped with a Microsoft Kinect and two Logitech QuickCam Pro 9000 cameras. These sensors allow the robot 3D vision and hearing. The head has an integrated 1 Watt audio amplifier and a speaker for speech output. Speech can be accompanied by an animation (e. g. of a mouth) on a graphic LCD. An adaptive gripping tool based on FESTO FinGripper fingers serves as a hand for the robot. The A1's torso is attached to a differentially steered mobile base that enables him to freely move around in his environment. So far I used the Kinect to let the robot follow a person with its gaze and to imitate a person’s movements. Furthermore I played around with voice output.
The project is work in progress. You can always find the latest info on the robot at my website www.malteahlers.de
When did you start working on this project and how long did it take you to complete?
In the middle of 2010 I started with pre-studies on the project (especially on the arms). About a year later I began building the current version of the robot. Since then I spent as much of my time as possible to further progress with the humanoid. Even if I already reached many important milestones of the project it is by far not finsehd yet.
What was your goal in building this project?
My ultimate aim is to build a universal humanoid robot that can autonomously operate in an environment made for humans and interact with humans in a human-like manner. Of course this is to ambitious to be reached to the full extent by a single developer. However, I think it is still worth to go in that direction as far as possible. Thereby one learns a lot about the problems of real world robotics. This self-educational aspect is actually one important motivation for the project. Apart from that, under certain "boundary conditions" the A1 might at some point actually be able to solve the classical "get me a beer" problem. Even if this would work only in a quite predefined environment it would still be an important achievement for me.
Does your project help to solve a problem? If so what problem?
Usually adult size humanoid robots are extremely expensive machines. I'm fascinated by the idea to build such a humanoid with (relatively) inexpensive standard components. As a hobbyist with restricted funds one can't use very expensive parts anyway. But furthermore I think that even in robotics research and industry it is a good strategy to begin to look for low cost solutions. Only by this means humanoids will become affordable to more people in the future. Of course I do not claim to have "solved" the problem of expensive humanoid robot. But if the A1 project further evolves as well as it did I see the possibility of making the project open source so that people can build their own humanoids on the basis of the A1 concept at modest cost.
What makes your idea unique?
The idea of building a humanoid robot is of course not unique at all. What maybe unusual to some degree is to actually develop an adult size humanoid as a hobby project.
In what capacity are you using bearings and what type of bearings?
Motion of some kind is essential for all robots. Therefore bearings are of high importance in robotics. Currently, the A1 humanoid has 17 axes whereof 10 are motorized. Different types of bearings are used for these axes. The stepper motors and planetary gearheads that drive the robot's arm and the mobile base as well as the servo motors of the sensor head contain ball bearings. The robolink arms and the two axes of the sensor head use plastic slide bearings. In the future a linear bearing should be added to the robot to make the torso vertically adjustable.
What is the most important thing you want people to know about your project?
I would feel happy if my project encourages other people to start robotic research and development on personal level. I'm convinced that anybody can have bright ideas for future technologies and especially for the field of robotics. Only if many people participate in these developments they will be useful for many people.
How will you use the $5,000 prize?
All self made parts for the humanoid were manufactured on a manual (non CNC) milling machine. For future work on the project it would be great to have at least a small CNC milling machine. Thus I would spend a large fraction of the prize for a solid CNC mill. Furthermore it would be great to have some money to take one or two month off to fully concentrate on the work on this project. At the moment this isn't possible for me.