A Practical Path Planner for the Robotic Vacuum Cleaner in Rectilinear Environments

In this paper, we propose a path planner for a robotic vacuum cleaner (RVC). In the design of the planner, we consider two main issues: (1) human-friendly path generation and (2) low computational load. First, we analyze how human move and suggest a hypothesis that human navigate in a way that minimizes the sum of muscle and brain energy. By imitating the human path, we propose a human- friendly path planner. Also, the designed planner requires a low amount of computations which not only extends the battery running time but also decreases the hardware cost of the RVC. Experimental results show that the proposed path gives a more favorable impression to customers than conventional paths.

I. Introduction

Nowadays, consumers pay considerable attention to Robotic Vacuum Cleaners (RVC). The market size had grown to 350 billion dollars in 2005 [1]. There are several problems to solve for ensuring the continual growth of the market. One problem is to have a natural motion when a RVC moves to other space (e.g., coming back to the charging station or going to other room to clean it up). For this natural motion, a localization system (which provides an accurate position of the RVC) and a path planner (which yields a path that gives a good impression to customers) have to be developed.