Vision-Guided Robot Companion

In the aging US. Population, conventional walking aids are widespread, and use is continuing to grow. Devices like a cane or rollator walker support the user's stability and prevent falls. While these devices are effective for their simplicity, they require the users to move the device manually. This operation can be physically and cognitively taxing as individuals concentrate on their mobility aid. To maintain stability without the burden of moving the device, a walking aid should move autonomously. This thesis presents the development of a robot companion that functions as a smart rollator walker. The robot has two modes of operation: standard and companion mode. In standard mode, the robot autonomously moves with the user by tracking the user's waist. The robot can maneuver in any direction, allowing for bidirectional movement and turning, like a conventional walker. In companion mode, the robot sits at the side of the user to assist with activities like shopping. The robot maintains its position at the side while the operator moves.The robot companion was evaluated through a series of tests for standard and companion mode. First, step response tests were completed for translational and rotational inputs. For companion mode these tests involved adding weights up to 30 lbs. The results of these experiments showed the robot's ability to precisely match the position of a user in a step response. The continuous ability to match the user was tested through a straight-line walking test. The results of these experiments showed the robot's ability to walk with the user over an interval of 25 seconds. Finally, system testing in an indoor environment was tested. This experiment was conducted in a real-life environment with a path involving long hallways and multiple turns. Results from the experiment validate the robot companion's ability to perform its intended functions in a longer test.

Grantor dc:publisher

University of Alabama Libraries

Year dc:date.issued

2024

Author dc:creator

• Fioresi, Braden

Advisor dc:contributor.advisor

• Shen, Xiangrong

Contributors dc:contributor

• Vikas, Vishesh
• Sazonov, Edward

• Controls
• LIDAR
• Raspberry-Pi
• Walking-Device

Statement dc:rights

• All rights reserved by the author unless otherwise indicated.

Language dc:language.iso

en_US, English

Dc Identifier Other

1102481

OAI identifier oai:identifier

oai:ir.ua.edu:123456789/15351