Supervised autonomy for AUVs using limited bandwidth communication channels

该项目旨在改善机器人平台与操作员之间的反馈联系，以提高水下勘测作业的有效性。在调查过程中，需要进行一定程度的调整，以使水下机器人能够对他们收集的数据做出反应。通常难以可靠地对自主系统进行编程以识别显著数据，特别是当使命涉及搜索其传感器签名可能难以先验确定的特定特征时。相比之下，人类通常擅长快速识别重要数据或确定何时使命没有实现其目标。该项目旨在开发新的声学通信方案，允许人类操作员和水下机器人之间的通信，利用机器学习，网络和通信理论的发展。

This project aims to improve the feedback link between robotic platforms and an operator, to increase the effectiveness of underwater survey operations. During surveys, some level of adaptation is required to allow underwater robots to respond to the data they are collecting. It is often difficult to reliably program an autonomous system to identify salient data, particularly when the mission involves searching for particular features whose sensor signatures may be difficult to determine a priori. In contrast, humans are generally good at quickly identifying important data or determining when a mission is not achieving its goals. The project aims to develop novel acoustic communication schemes that will allow communication between the human operator and the underwater robot, exploiting developments in machine learning, network and communication theory.