Models and Algorithms for Ultra-Wideband Localization in Single- and Multi-Robot Systems - Public Thesis Defense
Event details
| Date | 07.06.2013 |
| Hour | 17:00 › 18:30 |
| Speaker | Amanda Prorok |
| Location | |
| Category | Thesis defenses |
The United Nations predict that by 2020, over 80% of the worlds population will live in urban settings, with a large shift of activities to the indoors. Indoor localization is predictably a game-changing technology. In this context, this thesis focuses on the development of efficient methods for two complementary and orthogonal sub-domains of indoor localization, (i) ultra-wideband localization and (ii) collaborative localization, with the ultimate aim of presenting a unified framework, and understanding the benefits of their combined usage.
First, ultra-wideband (UWB) counts among the most interesting technologies for absolute indoor localization to date. Owing to its fine delay resolution and its ability to penetrate through various materials, UWB provides a potentially high ranging accuracy, even in non-line-of-sight (NLOS) environments. Our work improves upon state-of-the-art by addressing the peculiarities of UWB error behavior in harsh NLOS scenarios. By proposing a compact UWB error model and an efficient calibration algorithm, we are able to achieve centimeter-level accuracy in cluttered environments.
Second, collaborative strategies are particularly interesting because they can harness information distributed over large-scale networked systems. To this means, we develop a relative positioning model based on range and bearing measurements, and that allows easy integration with our UWB localization method. We tailor a unified localization framework that supports low cost, flexibility, and full decentralization, and that is a viable solution for large-scale systems composed of miniature devices. Finally, we take care to validate all aspects of our work with real world experimentation, using robot teams of up to 10 units in size.
First, ultra-wideband (UWB) counts among the most interesting technologies for absolute indoor localization to date. Owing to its fine delay resolution and its ability to penetrate through various materials, UWB provides a potentially high ranging accuracy, even in non-line-of-sight (NLOS) environments. Our work improves upon state-of-the-art by addressing the peculiarities of UWB error behavior in harsh NLOS scenarios. By proposing a compact UWB error model and an efficient calibration algorithm, we are able to achieve centimeter-level accuracy in cluttered environments.
Second, collaborative strategies are particularly interesting because they can harness information distributed over large-scale networked systems. To this means, we develop a relative positioning model based on range and bearing measurements, and that allows easy integration with our UWB localization method. We tailor a unified localization framework that supports low cost, flexibility, and full decentralization, and that is a viable solution for large-scale systems composed of miniature devices. Finally, we take care to validate all aspects of our work with real world experimentation, using robot teams of up to 10 units in size.
Practical information
- General public
- Free
Organizer
- Amanda Prorok