Humanoid navigation and heavy load transportation in a cluttered environment

Although in recent years several studies aimed at the navigation of robots in cluttered environments, just a few have addressed the problem of robots navigating while moving a large or heavy object. This is especially useful when transporting loads with variable weights and shapes without having to change the robot hardware. On one hand, a major advantage of using a humanoid robot to move an object is that it has arms to firmly grasp it and control it. On the other hand, humanoid robots tend to have higher drift than their wheeled counterparts as well as having significant lateral swing while walking, which propagates to anything they carry. In this work, we present algorithms for a humanoid robot navigating in a cluttered environment while pushing a cart-like object. In addition, the algorithms make use of the hands and arms to articulate the cart when executing tight turns using whole body control scheme to reduce the lateral swing effect on the load and ensure a safe transport. Experiments conducted on a real Nao robot assessed the proposed approach and algorithms, they show that the payload of a humanoid robot can be significantly increased without changing the humanoid robot's hardware, and therefore enact the capacity of humanoid robots in real-life situations.