Multi-Robot Tools and Techniques for Exploration and Mobile Surveying

无论是在地球上还是在其他地方工作，获取有关操作环境的准确、及时的地理空间信息对于规划、生产力和安全至关重要。例如，绘制岩体中的不连续面是设计采矿中适当的地面支持系统或了解火星上的地质过程的重要输入。然而，大多数当代地理空间测绘工具仍然是高度手动的，使用固定平台（例如，三脚架安装），因此在很大程度上依赖于人类操作员的物理存在和输入。大多数技术一次也只依赖于一个来源的数据，并依靠基于卫星的GPS信息来进行地理定位，这通常是不实际的（例如，在地下挖掘中）。这一应用程序的本质源于这样一种认识，即在将当代野外机器人的思想与偏远、恶劣、不确定和非结构化环境中的地理空间测绘的实际问题相结合时，存在着一个重要的、未被探索的机会，在这些环境中，人类的存在要么是不可能的，要么是以某种方式受到阻碍，要么是非常不受欢迎的（例如，地下深处，遥远的北方，在其他星球上）。这项研究方案的长期目标是发展“流动”测量和地理空间数据收集的工程基础，并将其应用于采矿和远程勘探活动；也就是说，创建新的可证明的、健壮的和实用的算法，这些算法比现有的技术有显著的改进，这些技术不利用移动设备的使用和/或从工程的角度来看缺乏发展。特别是，本研究侧重于将环境特征（例如，底层地质结构）与多个移动平台比固定测量设备提供的额外灵活性相结合的多机器人定位和测绘问题。该计划将资助完成硕士(3)和博士(2)的学生，通过两个短期研究方向发展相关技能（例如，系统建模，仿真和机器人技术），这些研究方向是在本资助申请所涵盖的时期内计划的：1)技术链将为移动机器人测量建立一个基本框架，该框架独立于可用的移动平台类型（例如，地面，空气，水），并尽可能独立于可用的传感器类型（例如，相机，激光雷达）；2)工具链将通过开发至少一个与采矿中岩体特征相关的实际示例来验证该研究。反过来，这项工作将有助于建立一个适用于一系列工程应用的基础研究体系，不仅包括行星和采矿业务，还可能包括土木工程和安全机器人。

Whether working on Earth or beyond, access to accurate and timely geospatial information about one's operating environment is critical to matters of planning, productivity, and safety.  For example, mapping of the discontinuity planes in a rock mass is an important input to the process of designing appropriate ground support systems in mining, or for understanding geological processes on Mars.  Yet, most contemporary tools for geospatial surveying and mapping remain highly manual, utilize stationary platforms (e.g., are tripod-mounted), and thus depend significantly on physical presence of and input by human operators.  Most techniques also rely on data from only one source at a time and on the availability of satellite-based GPS information for geolocation, which is often not practical (e.g., in underground excavations).  The essence of this application is born from the realization that there exists a significant and unexplored opportunity in the combination of ideas from contemporary field robotics with the real problems of geospatial surveying and mapping in remote, harsh, uncertain, and unstructured environments where human presence is either not possible, hindered in some way, or highly undesirable (e.g., deep underground, far north, on other planets).The long-term objective of this research program is to develop an engineering basis for "mobile" surveying and geospatial data collection, with applications in mining and remote exploration activities; i.e., to create new provable, robust, and practical algorithms that achieve marked improvement over existing techniques that do not exploit the use of mobile equipment and/or lack development from an engineering point of view.  In particular, this research focuses on multi-robot localization and mapping problems that combine features of the environment (e.g., underlying geological structures) with the additional flexibility that multiple mobile platforms afford over stationary surveying devices.  This program will fund to completion mostly Master's (3) and PhD (2) students to develop relevant skills (e.g., system modelling, simulation, and robotics) through two short-term research strands that are are planned for the period covered by this grant application: 1) a techniques strand will establish a fundamental framework for mobile robotic surveying that is independent of both the types of mobile platforms available (e.g., ground, air, water) and, as much as possible, independent of the types of sensors available (e.g., cameras, LiDAR); and, 2) a tools strand will focus on validating this research through the development of at least one practical example related to rock mass characterization in mining.  This work, in turn, will contribute to a body of basic research that is suitable to a spectrum of engineering applications, including not only planetary and mining operations, but possibly also civil engineering and security robotics.