Sensor Based Layered Control for Mobile Robots in Unknown Environments

Abstract

Existing robots are realized with embedded control. This makes it difficult to operate in unknown environments where environmental information cannot be obtained in advance. The objective of this research is to construct a control system for a robot that can reliably carry out its work objectives in an unknown environment. Implementing environmentally adaptive behavior requires a simultaneous discussion of the uncertainty of the environment perceived by the robot and the robot's chosen behavior. We propose a method to design robot behaviors according to sensor information and make this a single layer. The function to be achieved by the control system design at each layer is considered the work objective. The actions from multiple layers are stratified according to the sequential relationship of the work objectives. This paper describes the results of applying layered control based on sensor information to a wheeled robot and a quadrotor helicopter. Numerical simulations and experiments confirmed that target tracking can be achieved while avoiding obstacles. The next task is to construct a hierarchy of flying robots as an extension of wheeled robots.