Mahru

Mahru has a human-like body including a rotating head, arms, legs and six fingers plus three-dimensional vision to recognise chores that need to be tackled. The robot recognises people, can turn on microwave ovens, washing machines and toasters, and also pick up sandwiches, cups and whatever else it senses as objects.

Korea Institute of Science and Technology (KIST)

Specifications

Mahru III
Height 150 cm
Weight 62 kg
Degrees of freedom32
Maximum walking speed (Current)1.3 km/h
Maximum walking speed (Target)2.6 km/h
Maximum height of bump1 cm
Battery operation hour30 m

Overview

The robot recognises people, can turn on microwave ovens, washing machines and toasters, and also pick up sandwiches, cups and whatever else it senses as objects. The idea behind this prototype is to get it to repeat the movements of a human by mimicry. Somebody moves in a certain way and the robot moves in exactly the same manner, acting as a remote-controlled telepresence robot.
The user wears an Xsens MVN suit while performing various tasks. The movements are recorded and the robot is then programmed to reproduce the tasks while adapting to changes in the space, such as a displaced object. 


The key feature of this humanoid, are that the mechanical system is carefully designed by CAE (Computer Aided Engineering) and
DOE (Design of Experiments)-based optimization. The purpose of this optimal design process is the realization of lightweight, high stiffness, and high energy efficiency. Toward these goals, kinematic structure is at first designed to minimize load torque guaranteeing large motor speed limits of the lower body, and to maximize manipulability measure of the upper body. Then, CAE analysis and topology optimization for metallic frames are done aiming at light weight and high stiffness of the whole mechanism. Actuating systems are selected to minimize motor power consumptions.

References

Presents the system overview, the design of mechanical systems and electrical/electronic parts, and the outline of the walking control algorithm. Describes the experimental results of the current phase and future research targets for improved motion performance and network-based applications.

Woong Kwon, J.A.Park, Hyun K. Kim, et al. - 2008

Describes a wearable-type motion capturing system, its design and implementation.

HeeBae Song, Doik Kim, Mignon Park, et al. - 2008

Introduces the architecture for overall control system including an internal control system and an external control system. Shows dynamic walking mechanism for the humanoid. Introduces network-based intelligence to recognize human beings, gesture, voice, three-dimensional objects. Shows the develope

Bum-Jae You, Changhwan Kim, Doik Kim, et al. - 2008

Wevolver 2023