Converting Natural Language Text to ROS-Compatible Instruction Base

Abstract

Natural Language processing is a growing field. Although it is difficult to create a natural language system that can robustly react to and handle every situation, it is quite possible to design the system to react to specific instructions or scenario. The contributions of this work are (1) to design a set of instruction types that can allow for conditional statements within natural language instructions, (2) to create a modular system using Robot Operating System (ROS) in order to allow for more robust communication and integration, and (3) to allow for an interconnection between the written text and derived instructions that will make the sentence construction more seamless and natural for the user. As the results will show, this system can be run on a diverse set of sentence structures, allowing for robust paragraphs. This system must also then be carefully set to fit the exact parameters that the user is looking for, trying to strike the balance between how much the user needs to learn and how accurate to the instruction the system needs to run.