CSR-EHS: Collaborative Research: A General, Efficient and Robust Platform for Enabling Control Applications in Sensor Networks

本研究项目的目标是建立一个应用基础设施和一套嵌入式系统软件，使软件控制和监测应用，如加热，通风和空调系统（HVAC），在低功耗，有损无线传感器-执行器网络的快速发展。这种用户偏好驱动的控制应用需要能够适应大规模分布式网络中的噪声、损失和故障，并且需要在获取用户偏好、做出控制决策并将这些决策传播给适当的执行器方面实现高度自动化。正在开发的传感器控制系统（SCS）提供了一个高级平台，允许用户在整个网络的粒度上表达控制应用需求，使用户能够通过抽象低级传感器网络细节来关注其部署的需求。SCS将机器学习和统计技术与软件系统和数据库中受启发的高级抽象相结合，允许在这些系统中固有的损失和不确定性的情况下进行格式合理的决策。该项目将在高度仪器化的办公环境中演示该方法。这项研究有可能通过开发一类新的无线控制系统来影响工业，也有可能通过将跨学科的思想注入到EECS的各个子领域来影响学术界。在本研究过程中开发的数据集、代码和传感器执行器部署将用于一系列教育计划。

The goal of this research project is to build an applicationinfrastructure and a suite of embedded system software that willenable the rapid development of software control and monitoringapplications, such as heating, ventilation, and air conditioningsystems (HVAC), on low-power, lossy wireless sensor-actuatornetworks. Such user-preference driven control applications requirethe ability to adapt to noise, loss, and failures in a large-scaledistributed network, and demand a high degree of automation inacquiring user preferences, making control decisions anddisseminating those decisions to the appropriate actuators.The Sensor Control System (SCS) being developed provides a high-level platform that allows users to express control applicationrequirements at the granularity of the entire network, enabling usersto focus on the requirements of their deployments by abstracting awaythe low-level sensor network details. SCS combines techniques frommachine learning and statistics with high-level abstractions inspiredby work in software systems and databases, allowing formathematically-sound decision making despite loss and uncertaintythat is inherent in such systems. The project will demonstrate the approach ona highly instrumented office environment. This research has the potential toimpact industry through the development of a new class of wirelesscontrol systems, and academia through the infusion of cross-disciplinary ideas into a variety of sub-fields of EECS. Thedatasets, code and the sensor-actuator deployments developed duringthe course of this research will be used in a range of educationalinitiatives.