We introduce the concept of architectural "component-sharing" as the basis for "knowledge sharing" in hive minds, e.g., a system multiple robots connecting by shared components in their control architectures. We discuss the architectural requirements and demonstrate the utility for multi-robot instruction and automatic reasoning across robotic plat- forms with two examples of natural language human-robot interactions with mind-sharing robots. Finally, we discuss some of the challenges of making instructing hive minds intuitive for humans and point to ques- tions that need to be addressed in the future.
@inproceedings{scheutz21icai,
title={Mind Readers and Mind Users: The Utility of Sharing Architectural Components across Multiple Robots},
author={Matthias Scheutz},
year={2021},
booktitle={23rd International Conference on Artificial Intelligence},
url={https://hrilab.tufts.edu/publications/scheutz21icai.pdf}
}