Robot simulation experiments

PhD student, Nadine Wirkuttis and Prof. Tani set up three experiments looking at robot interactions where they introduced an additional parameter, w, into the robot’s learning process. The results revealed that w, referred to as the meta-prior, determined whether the robots are more likely to be driven by their own intention or by sensory observation. When the w was set high, the robot only generated its own patterns as it intended, whereas when the w was set low, the robot only followed what it observed in the counterpart. The researchers found that choosing the same w value for both robots in an adequate range allowed the robots to take turns between leading and following spontaneously.

PhD student, Nadine Wirkuttis and Prof. Tani set up three experiments looking at robot interactions where they introduced an additional parameter, w, into the robot’s learning process. The results revealed that w, referred to as the meta-prior, determined whether the robots are more likely to be driven by their own intention or by sensory observation. When the w was set high, the robot only generated its own patterns as it intended, whereas when the w was set low, the robot only followed what it observed in the counterpart. The researchers found that choosing the same w value for both robots in an adequate range allowed the robots to take turns between leading and following spontaneously.

This video appeared in the news article "Expectations vs reality: How robots update their behavior based on their experience".

Date:

12 December 2021

Creator:

lucy-dickie

Copyright OIST (Okinawa Institute of Science and Technology Graduate University, 沖縄科学技術大学院大学). Creative Commons Attribution 4.0 International License (CC BY 4.0).

Tags

Research

Researchers

Share on:

Related Videos

To Buckle or Not to Buckle

Testing the stability of cylindrical tube made from 15,360 hexagonally packed magnetic balls (192 rings, each made up of 80 balls). A high-speed camera recording reveals crumpling at the base of the cylindrical tube, reminiscent of paper-like crumpling, when it loses stability from shaking.

The mechanophore glows more intensely as the stretching increases.

The polymer material is stretched with an increasing force resulting in a corresponding brighter emission of light from the mechanophore (here under UV light and with false colors). The graph shows the intensity of the emitted light following punctual increases in the stretching force applied to the polymer.

Fracton Models

PhD Student Han Yan in the Theory of Quantum Matter Unit at the Okinawa Institute of Science and Technology Graduate University (OIST) put a fracton model in a curved space – where holography works for gravity. Yan checked if the properties also work for fracton models, and they did.

Math Meets Art: Möbius Kaleidocycles

Dr. Johannes Schönke and Prof. Eliot Fried of the OIST Mathematics, Mechanics, and Materials Unit have introduced a new class of kaleidocycles into the world. They call them Möbius Kaleidocycles because they resemble the famous Möbius band, a geometrical object with a characteristic topology. These mystifying objects can be turned inside-out continuously and have unique mathematical properties. While classical kaleidocycles typically have six hinges, the new class of kaleidocycles can have seven or more. Not only are these objects beautiful to see, but they could also have very practical applications in a variety of fields.

Fluorescent Calcium Imaging

The neurons in this video contain a dye that fluoresces when it encounters calcium ions. When the neuron is excited, calcium floods into the cell and the neuron fluoresces. The researchers monitored experiments such as this one to see which neurons were firing synchronously and which were firing separately.