Date of Award:


Document Type:


Degree Name:

Doctor of Philosophy (PhD)


Computer Science

Committee Chair(s)

Mario Harper


Mario Harper


Xiaojun Qi


Curtis Dyreson


John Edwards


Jacob Gunther


Exploring unknown environments using multiple robots poses a complex challenge, particularly in situations where communication between robots is either impossible or limited. Existing exploration techniques exhibit research gaps due to unrealistic communication assumptions or the computational complexities associated with exploration strategies in unfamiliar domains. In our investigation of multi-robot exploration in unknown areas, we employed various exploration and coordination techniques, evaluating their performance in terms of robustness and efficiency across different levels of environmental complexity.

Our research is centered on optimizing the exploration process through strategic agent distribution. We initially address the challenge of city roadway coverage, aiming to minimize the travel distance of each agent in a scenario involving multiple agents to enhance overall system efficiency. To achieve this, we partition the city into subregions. and utilize Voronoi relaxation to optimize the size of postman distances for these subregions. This technique highlights the essential elements of an efficient city exploration.

Expanding our exploration techniques to unknown buildings, we develop strategies tailored to this specific domain. After a careful evaluation of various exploration techniques, we introduce another goal selection strategy, Unknown Closest. This strategy combines the advantages of a greedy approach with the improved dispersal of agents, achieved through the randomization effect of a larger goal set.

We further assess the exploration techniques in environments with restricted communication, presenting upper coordination mechanisms such as frontier incentivization and area segmentation. These methods enhance exploration performance by promoting independence and implicit coordination among agents. Our simulations demonstrate the successful application of these techniques in various complexity of interiors.

In summary, this dissertation offers solutions for multi-robot exploration in unknown domains, paving the way for more efficient, cost-effective, and adaptable exploration strategies. Our findings have significant implications for various fields, ranging from autonomous city-wide monitoring to the exploration of hazardous interiors, where time-efficient exploration is crucial.



Creative Commons License

Creative Commons Attribution-Noncommercial 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 4.0 License.