Adaptive Systems for Collaboration in Multi-mode Mobile Environments

这项研究将沿着两个方面推动群件技术的发展，这两个维度都是由移动性驱动的：(1)它将根据参与者提供的策略和指导方针以及响应于组配置的变化来自动化群件任务的管理；这一过程将通过开发适当的形式化、协议和度量来实现。(2)它将开发一种软件基础设施，为动态无线环境中的群件量身定做，并能够进行多模式操作，即，在通过网关连接到基础设施网络时以游牧方式运行，或在没有外部资源可用时以临时方式运行。该方法的核心是知识管理的概念，即环境、其他设备和软件功能的知识由系统内的不同移动设备收集和交易，总体目标是增加每个移动设备对其感知的物理和计算环境的了解。这项工作的关键技术贡献是将群件引入移动自组织网络领域，并实现基本群件过程的自动化。无线自组织网络的动态化使得现有的群件技术无法提供简单的端口，以适应可靠的有线系统。事实上，需要重新审视群件的概念，需要考虑适应移动性的新算法和协议。群件过程的自动化也带来了其自身的挑战，因为复杂的任务、动作、信息流、工作流、应急计划等需要以机器可读的格式来描述。此外，移动网络中一直存在的不确定性因素进一步增加了开发可靠算法和指标的挑战。群件包含一系列促进人类协作的软件技术，从简单的电子邮件和公告牌服务到共享白板和视频会议。今天，群件技术正处于一场革命的风口浪尖，这场革命主要是由移动设备技术令人难以置信的发展速度推动的，移动设备技术已经将蜂窝电话和PDA转变为可行的便携式计算机，进而成为群件的下一代平台。将移动性与群件技术相结合，将为雄心勃勃的新应用程序奠定基础，这些应用程序将促进和管理人群的协调行为。应对化学品泄漏等灾难，对偏远地区进行地质调查，甚至管理探索偏远星球的机器人社区，这些只是许多复杂、高度协调的活动中的几个例子，这些活动将从这种方法中受益。该项目还将指导博士生，引导本科生进入博士项目，并通过与圣路易斯科学中心的持续合作努力促进公众对科学和技术的理解。

This research will advance the state of the art in groupware along two dimensions, both motivated by mobility: (1) It will automate management of groupware tasks, subject to policies and guidelines provided by the participants and in response to changes in group configuration; the process will be made possible by the development of appropriate formalizations, protocols, and metrics. (2) It will develop a software infrastructure tailored to groupware in dynamic wireless environments and capable of multi-mode operation, i.e., in a nomadic manner when connected to infrastructure networks through a gateway or in an ad hoc manner when there are no external resources available. Central to the approach is the concept of knowledge management, whereby knowledge of the environment, other devices, and software capabilities is collected and traded by different mobile devices within the system with the overall aim of increasing each mobile device's understanding of the physical and computational environment that it perceives. This knowledge can then be used to enhance the capabilities of each entity within the system, and therefore by extension the collaborative behavior supported by the system itself.The key technical contribution of this work is the introduction of groupware into the realm of mobile ad hoc networks and the automation of fundamental groupware processes. The dynamism of wireless ad hoc networks makes infeasible simple ports of current groupware technology, geared for reliable wired systems. In fact, the very notion of groupware needs to be reexamined and new algorithms and protocols that accommodate mobility need to be considered. Automation of groupware processes also introduces its own challenges since complex tasks, actions, information flows, work flows, contingency plans and much more need to be described in a machine decipherable format. Additionally, the element of uncertainty, ever present in mobile networks, further adds to the challenge of developing reliable algorithms and metrics. Groupware encompasses a broad range of software technologies that facilitate human collaboration, ranging from simple e-mail and bulletin board services to shared whiteboards and videoconferencing. Today, groupware technology is on the cusp of a revolution, one that is being driven primarily by the incredible pace of development of mobile device technology that has transformed cellular phones and PDAs into viable portable computers and, by extension, into next generation platforms for groupware. Blending mobility with groupware technology will establish a foundation for ambitious new applications that will facilitate and manage the coordinated behavior of groups of people. Responding to catastrophes such as chemical spills, conducting geological surveys of remote areas, and even managing a community of robots exploring a remote planet are only a few examples of the many complex, highly coordinated activities that will benefit from this approach. The project will also mentor doctoral students, channel undergraduates towards doctoral programs, and contribute to public understanding of science and technology through an ongoing collaborative effort with the St. Louis Science Center.