Driving by Speaking: Natural Language Control of Robotic Wheelchairs

While many people use electric wheelchairs to enhance their mobility, there are groups who cannot due to age, injury, or cognitive impairment. A voice-operated smart wheelchair would greatly enhance the self-mobility of these users. This thesis presents a method to parse natural language commands into useful robotic instructions as well as control methods to navigate in a priori known and novel environments. An agglomerative algorithm was developed that uses stemmatic processing to extract meaning from natural language commands. A method for creating a navigable graph from a simplified blueprint is introduced. A MATLAB simulation was created to simulate navigation in a novel building using sensory data obtained from a virtual LIDAR. The natural language navigation control algorithms have been tested for destination-based and landmark-based navigation and perform well on simple commands.

Name thesis:degree_name

Master of Sciences (Engineering)

Level thesis:degree_level

masters

Discipline thesis:degree_discipline

EECS - Electrical Engineering

Grantor dc:publisher

Case Western Reserve University School of Graduate Studies

Year dc:date

2013

Author dc:creator

• Hecht, Steven A.

Contributors dc:contributor

• Lee, Gregory

• Robotics
• Mobile Robotics
• Cognitive Robotics
• Natural Language Processing
• Navigation
• Simulation

Statement dc:rights

• unrestricted
• This thesis or dissertation is protected by copyright: all rights reserved. It may not be copied or redistributed beyond the terms of applicable copyright laws.

Language dc:language

English

Repository record dc:identifier

http://rave.ohiolink.edu/etdc/view?acc_num=case1365158700

OAI identifier oai:identifier

oai:etd.ohiolink.edu:case1365158700