humanoid multimodal tactile sensors hd

Loading...

Sign in or sign up now!
Alert icon
Upgrade to the latest Flash Player for improved playback performance. Upgrade now or more info.
28,349
Loading...
Alert icon
Sign in or sign up now!
Alert icon

Uploaded by on Jan 25, 2011

Part I - Tactile Module:
Time: 00:00-00:09
------------------------------
In the first part we show the newly developed Tactile Module in comparison with a human hand.

On the backside the local controller is visible next to the four combined power and data ports. One BMA150 3-axis accelerometer (black box middle), six PT1000 temperature sensors (blue boxes) and four GP2S60 proximity sensors (outer black boxes) are visible on the front side.

Part II - Robot reaction towards different modalities:
Time: 00:10-00:29
---------------------------------------------------------------------
In the second part we show reactions of a KUKA robotic arm towards the different modalities of a single sensor - proximity, acceleration and temperature.

The robot was programmed to evade on a change of the according sensor value and return to his home position otherwise. With the proximity and temperature experiments only an evasive reaction was programmed. With the impact detection experiment we also implemented an inversion of the reaction, this is why the robot stays within the new boundaries. Please note that there is only one sensor mounted on the visible side of the robot. In general the control loop of these experiments is running at 1kHz and we only actuate the base joint.

Part III - Robot reaction towards multiple modules:
Time: 00:30-01:02
-------------------------------------------------------------------
In the last part we show the performance of a network of connected modules, driving multiple inverse kinematic chains of the KUKA robot.

In this part we always show two states of the experiment, one with our tactile controller turned off, the other one with the controller turned on. Please note that KUKA light weight robotic arm has impedance control capabilities which is used in the orientation experiment to push the robot around.
In the first experiment we show our multitouch controller reacting to proximity sensor input. In contrary to a purely joint force controlled robot we can detect the location of touch if multiple forces on one segment equalize and interact without applying any force.
In the second experiment we show the stabilization of a segment equipped with our sensor module towards a superposition of movement acceleration and gravity. This helps for example to stabilize a loosely placed cup on a tablet. In contrary to other possible solutions we do not need a fixed base nor a very accurate kinematic model. In general the control loop of these experiments is running at 0.1 kHz and we use an inverse kinematic chain per segment. The joint excitation of every inverse kinematic chain is summed up and sent to the robot.

P. Mittendorfer and G. Cheng, "Humanoid multi-modal tactile sensing modules," IEEE Transactions on Robotics, vol. 27, no. 3, pp. 401--410, June 2011.

Category:

Science & Technology

Tags:

License:

Standard YouTube License

Link to this comment:

Share to:

Top Comments

  • FANTASTIC. The Handjob Robot is almost complete.

see all

All Comments (10)

Sign In or Sign Up now to post a comment!
  • it is even better than human skin!!! the infrared sensors work like invisible wiskers!

    you made my day too! it is soooo smooth in motion!!!! i need to show it to my girlfriend to shock her too! keep it up, great piece of technology!!!

  • Und dann? World War 3?

  • Looks great! Well done. Similar work also demonstrated on the video session of ICRA 1999

  • Robots with feelings? lol

  • Seems to work quite nicely with the arm, although I really want to see this built into a machine capable of motion and handling. Something with more limbs, like a humanoid or insectoid robot.

    It'd be quite interesting to see such a robot navigate rough environments.

  • Have they mapped out the human brain including all the neurons yet? 

  • Actually, with just the demonstration, it's obvious you can use a robot arm to move heavy objections by directly interacting with the arm which is more intuitive as opposed to a joy stick remote.

  • @EstanBulLoFre you really made my day.... lol

  • AMAZING!!!

    

Loading...

Alert icon
0 / 00Unsaved Playlist Return to active list
    1. Your queue is empty. Add videos to your queue using this button:
      or sign in to load a different list.
    Loading...Loading...Saving...
    • Clear all videos from this list
    • Learn more