Fault-tolerant distributed algorithms and realistic models for groups of autonomous mobile robots

本研究旨在促进容错可靠多机器人系统的发展。特别是，这包括将容错分布式系统领域中开发的形式化模型与移动机器人协调问题联系起来。这种模型的目的是研究算法的正确性，并确定单个机器人为了可靠地解决给定问题而必须具备的最低能力(即，传感器集)。同时，模型和问题的精确度也是至关重要的，这样才能使结果直接适用于实际系统。特别是，我们在几个问题上得到了重要的结果，其中包括：(1)当一些机器人或它们的传感器可能发生故障时机器人的聚集；(2)离散环境的探索；(3)多机器人的运动规划；(4)大系统中的连锁故障。

This research aims at improving the development of fault-tolerant and reliable multi-robots systems. In particular, this includes relating formal models developed in the field of fault-tolerant distributed systems to the problem of mobile robot coordination. The purpose of such a model is to study the correctness of algorithms and identify the minimum capabilities (i.e., set of sensors) that the individual robots must have in order to reliably solve a given problem as a group. At the same time, it is essential that both model and problem be practically accurate so that results are directly applicable to real systems.In particular, we have obtained important results for several problems, among which,(1) gathering of robots when some robots or their sensors may fail, (2) exploration of a discrete environment, (3) motion planning of multiple-robots, and (4) cascading failures in large systems.