#glc2016

Research Theme: SLAM on Humanoid Robot Platform

1. Reasech Progress

1.1. Background and Motivation

1.2. Introduction - Odometry and SLAM

Odometry is changes in translation and rotation of robots. It is estimated from a difference in sensor wheel or IMU data between time t and t-1. 

Consecutive odometry data present a map.

Here is a created map by 2D SLAM.

Visual odometry is a common approach in 3D SLAM. Instead of laser range data, camera data are used. Visual odometry is estimated from a difference between time t and t-1.

Estimated translation along X axis.

Estimated translation along Y axis.

Estimated translation along Z axis.

Here is a created map by 3D SLAM.

1.3. Research Progress in 3D SLAM with Humanoid Locomotion

Applying a existing 3D SLAM algorithm RTAB-Map, our humanoid estimated odometry data during locomotion and created a 3D map.

2.1. Research Plan - Brushing up 3D SLAM with Humanoid Locomotion

Here is my future plan.

2.2 Research Plan - Internship

In my internship, I’m planning to participate in NASA Space Robotics Challenge. Here is what we do in the challenge.

2.3 Research Progress & Plan - Time Line

2.3.1 Progress

2.3.2 Plan