Towards self-driving navigation control systems for a rideable ballbot

Mansouri, Mahshid

Permalink

https://hdl.handle.net/2142/129898 Copy

Description

Title

Towards self-driving navigation control systems for a rideable ballbot

Author(s)

Mansouri, Mahshid

Issue Date

2025-05-30

Director of Research (if dissertation) or Advisor (if thesis)

• Hsiao-Wecksler, Elizabeth T.
• Driggs-Campbell, Katie R.

Doctoral Committee Chair(s)

• Hsiao-Wecksler, Elizabeth T.
• Driggs-Campbell, Katie R.

Committee Member(s)

• Amato , Nancy M.
• Norris , William R.
• Ramos, Joao

Department of Study

Mechanical Sci & Engineering

Discipline

Mechanical Engineering

Degree Granting Institution

University of Illinois Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

Ballbot, Dynamically-stable Robot, Autonomous Navigation, Path Planning, Obstacle Avoidance, Crowd Navigation, Human-robot Interaction, Mobility Device, Assistive Technology.

Abstract

Ballbots, a class of dynamically balancing mobile robots that use a single spherical wheel for omnidirectional locomotion, offer agile maneuverability and a compact footprint, making them well-suited for navigation in crowded or confined indoor environments. While prior work has explored path planning for ballbots, few have addressed the challenges of navigating with real-world payload constraints, and none have tackled autonomous navigation on a rideable ballbot that includes a human in the control loop. This dissertation presents the implementation and evaluation of autonomous navigation algorithms for MiaPURE (Modular, Interactive, Adaptive, Personalized, Unique Rolling Experience), a human-ridable, hands-free, dynamically balancing ballbot platform developed by our research group at University of Illinois at Urbana-Champaign (Human Dynamics and Controls Lab). This dissertation spans three investigations. First, we integrated and benchmarked a path planning technique on MiaPURE as a heavy-payload ballbot, demonstrating its performance across static and single dynamic obstacle environments at various payloads and speeds. Second, we extended this architecture to include a reactive local planner capable of predicting and avoiding multiple moving agents using Kalman filter-based pedestrian tracking, enabling crowd-aware navigation in low to moderately dense environments. Finally, we evaluated MiaPURE in rideable scenarios with human subjects, examining how different levels of autonomy, including driver control, shared control, and full autonomy, affect user performance, cognitive load, and safety. Together, these studies contribute to the advancement of real-world ballbot navigation by addressing challenges related to underactuated dynamics, obstacle avoidance in human-centric environments, and the integration of human-robot shared control. The findings lay the groundwork for developing future assistive mobility systems that combine dynamic stability, intelligent navigation, and user-centered interaction.

Graduation Semester

2025-08

Type of Resource

Thesis

Handle URL

https://hdl.handle.net/2142/129898

Copyright and License Information

Copyright 2025 Mahshid Mansouri

Owning Collections