igus open source soccer robot (with University of Bonn), first kicks…

We have been involved in the development of a new open source teen size soccer robot for 2 years now (and the improvement still goes on!). The main development was done by the University of Bonn, institute AIS, Professor Sven Behnke, who also developped our open source robolink software IME (igus motion editor), igus contrubuted the design and technology of the outer shell parts (made by 3D printing / SLS). In the next stage we will improve the cabeling inside the robot with our chainflex know-how for often bending cables.

This youtube video shows the current state of the software development (thanks to Philipp Allgeuer for sharing it).

The open source project is documented by Prof. Behnke here with open source data: http://www.nimbro.net/OP

And the 3D Step data of the robot shell (3D printed parts) are shared here by igus: https://github.com/igusGmbH/HumanoidOpenPlatform

igus open source soccer robot (with University of Bonn), first kicks...

robot FLASH from Wroclaw University of technology (Poland)

The team of Jan Kedzierski (Ph.D.) from Wroclaw University of technology now presents their NEW generation of the autonomous robot FLASH. It is (again) equipped with 2 pcs. 5 DOF robolink wire driven joint arms. They developpped their own drive technology for this robot based on DC/BLDC motors as an open source project. It will be published soon, I will link to their project then.


robolink D test area

we do several lifetime and performance tests with our robolink D joints. Until July, main topics were material and geometry optimisations for the gear and bearing components. Nowadays, we do lifetime tests for different joint types at different loads and speeds. The attached film shows or test unit with currently 9 test places, it will be enlarged in the next weeks up to 12 oor 14 test places.


Some numbers and facts that we found out until now:

# general efficiency factor for our current material combination (iglidur J with aluminum HC) is betweeen 0,28 and 0,35 (over all sizes).

# output torque at tooth-break: 19 Nm (size -20) and 30 Nm (size -30), measured at room temperature for first movement. Today we measure this “breaking torque” at continuous operation. Main influences are the rotation speed, and together with this, the temperature in the gear. We see a huge loss of “breaking torque” in operation. For size -20 we find a safe number at high speed rotation of 2,5 Nm (>100.000 cycles) and 3,5 Nm at low speed rotation.

we see good potential for improvement in the material combination and geometry optimisation.

RL-D different versions, pricing

We have defined 3 different quality standards for our robolink D joints and I want to explain the background of this issue. The 3 qualities are:

  • “low cost”,
  • “standard”,
  • “high end”

The main performance differences are stiffness, friction, precision (reverse play) and on the other hand cost. For size -20 and -30, the price difference between low cost and high end is ~factor 2, for size -50 it is ~factor 1,5. standard version ~ in the middle.

The high end version contains of our lubrication free worm gear with worm wheel made of iglidur J and worm shaft made of aluminum (hard anodized), it has ball bearings for the axial shaft load and for the back side (only version 102). The outer bearings are our PRT-01-xx (= high end rotary table made of aluminum with sliding elements made of iglidur J). Housing and inner connection parts are made of moulded plastic parts (our RN33), screws and nuts are stainless steel.

Most parts of the standard version are the same, but the back side has a sliding bearing (iglidur J) instead of the ball bearing (only version 102) and the PRT itsself is version PRT-02-xx-AL, which is a low cost version of PRT-01-xx. It has more plastic parts and thus is much lighter. Compared with PRT-01 is is a bit less stiff and a bit less precise. In a performanmce comparison of the to joints, we see no big differences for most applicaations as long as we do not get to the limits.

The low cost version of our RL-D in comparison contains a moulded plastic worm shaft and sliding axial bearings instead of the ball bearings. Also the PRT is not build with an aluminum outer ring but a moulded plastic outer ring (PRT-02-xx-LC). Friction is higher and precision worse than our “standard” version. It should be used only for “easy applications” as for example hand adjustments.  



For size -20 and -30, all versions are available. For size -50, only version “high end” is available, as PRT-02-50 is yet under construction. We will have it in ~OCT 2015.



robolink D data sheets and STEP data

Please find attached current (preliminary) data sheets for our new robolink D Joints. These Joints exist in 3 sizes: -20, -30 and -50 (based on the Dimensions of our rotary sliding tables PRT).

RL-D-xx-101 RL-D-xx-102

RL-D-xx-101: symmetrical Joint with 2 PPRT for vertical Installation (used in our robotic arms),

tech data RL-D-101 _D-ENG_.pdf (167.44 kb)

RL-D-xx-102: asymmetrical Joint with 1 PRT for horizontal Installation (used as positioning device or as 1st axis in a robotic arm)

tech data RL-D-102 _D-ENG_.pdf (150.10 kb) 

I will publish 3D Step data of These Joints beginning of next week in this blog.