Off-campus UMass Amherst users: To download campus access dissertations, please use the following link to log into our proxy server with your UMass Amherst user name and password.

Non-UMass Amherst users: Please talk to your librarian about requesting this dissertation through interlibrary loan.

Dissertations that have an embargo placed on them will not be available to anyone until the embargo expires.

Author ORCID Identifier


Open Access Dissertation

Document Type


Degree Name

Doctor of Philosophy (PhD)

Degree Program

Computer Science

Year Degree Awarded


Month Degree Awarded


First Advisor

Shlomo Zilberstein

Subject Categories

Artificial Intelligence and Robotics


Metareasoning is the process by which an autonomous system optimizes, specifically monitors and controls, its own planning and execution processes in order to operate more effectively in its environment. As autonomous systems rapidly grow in sophistication and autonomy, the need for metareasoning has become critical for efficient and reliable operation in noisy, stochastic, unstructured domains for long periods of time. This is due to the uncertainty over the limitations of their reasoning capabilities and the range of their potential circumstances. However, despite considerable progress in metareasoning as a whole over the last thirty years, work on metareasoning for planning relies on several assumptions that diminish its accuracy and practical utility in autonomous systems that operate in the real world while work on metareasoning for execution has not seen much attention yet. This dissertation therefore proposes more effective metareasoning for planning while expanding the scope of metareasoning to execution to improve the efficiency of planning and the reliability of execution in autonomous systems. In particular, we offer a two-pronged framework that introduces metareasoning for efficient planning and reliable execution in autonomous systems. We begin by proposing two forms of metareasoning for efficient planning: (1) a method that determines when to interrupt an anytime algorithm and act on the current solution by using online performance prediction and (2) a method that tunes the hyperparameters of the anytime algorithm at runtime by using deep reinforcement learning. We then propose two forms of metareasoning for reliable execution: (3) a method that recovers from exceptions that can be encountered during operation by using belief space planning and (4) a method that maintains and restores safety during operation by using probabilistic planning.


Creative Commons License

Creative Commons Attribution 4.0 License
This work is licensed under a Creative Commons Attribution 4.0 License.