Autonomous Planning and Mapping for the Characterization of Gamma Contaminated Environments
Loading...
Authors
Mascarich, Frank
Issue Date
2021
Type
Dissertation
Language
Keywords
Aerial Robotics , Planning , Radiation , Scalar Field Mapping
Alternative Title
Abstract
The past $100$ years of research and development in the fields of nuclear power, weapons, and industrial radiation applications have imbibed regions across the world with facilities and terrain which is contaminated with radioactive material. Such locations can pose significant hazards to human health, thus requiring vigilant monitoring and mitigation efforts. The use of autonomous robots is well suited to this task. Motivated by this fact, this work contributes a holistic perspective on the deployment, design, and use of autonomous robots for the characterization of radioactively contaminated environments. The set of developments presented in this dissertation incorporate principles of gamma radiation detection and measurement, techniques for mapping and localizing a variety of radioactive sources, path planning strategies tailored to both ground and aerial platforms, as well as prototype systems implementing methods for perception and navigation in dirty, dangerous, and degraded conditions. Specifically, Chapter \ref{chap:intro} presents the motivation behind this work, including its practical application, as well as a brief description of the approach utilized to accomplish environmental radiation characterization. Chapter \ref{chap:contrib} presents a detailed overview of the presented radiation mapping contributions and associated publications in addition to a brief note on other synergistic contributions made towards enabling autonomy in the perceptually degraded environments associated in particular with waste decommissioning facilities. Subsequently the core contributions of this thesis are presented in detail. Chapter \ref{chap:single_source} presents a method for autonomous single source localization using an aerial robot, alongside details regarding principles of radiation measurement and detection. Chapter \ref{chap:radbot} describes a technique developed to map distributed radiation fields in 2D using a ground platform, while Chapter \ref{chap:radmf} extends the work to perform the mapping task in 3D using a collision tolerant micro aerial vehicle. Subsequently, Chapter \ref{chap:auro} presents autonomous distributed 3D radiation mapping coupled with an intelligent path planning algorithm tailored to source seeking behaviors in confined environments. Finally, conclusions and an outlook for future research are discussed in Chapter \ref{chap:conclusions}.Overall, this dissertation contributes a body of work enabling autonomous radiological surveying in challenging conditions, demonstrating robust functionality through a series of field experiments using real radiation sources. Each of the presented methods is associated with a tested and reliable robotic system purpose-built for its designated task. This combination of performance robotic hardware demonstrating novel autonomous functionality in realistic use-case scenarios showcases the applicability and dependability of the presented systems and methods.