Analytics platform for correlated sensor information

物联网（IoT）和智能设备的最新进展导致IoT行业快速增长，预计到2020年全球市场将达到8.9T美元（http：//i.sfu.ca/TgFkjR）。该行业跨越了各个领域 **，从家用个人设备到工业级大型设备的监控和管理。** Plasmatic Technologies Inc.是一家总部位于温哥华的加拿大初创企业，致力于让具有前瞻性的 ** 组织及其客户从互联家庭设备以及他们通过家庭洞察和机器学习产生的有价值数据 ** 中受益。该公司独特的Alana智能家居平台和 ** 应用程序为创新的服务提供商计划提供支持，这些计划可以实现差异化、更强的客户参与度、收入增长和运营效率。Alana智能家居平台和应用程序旨在让客户无缝连接许多不同的智能家居技术，为他们提供洞察力和控制力，同时允许他们灵活地使用他们想要的设备。智能家居系统和互联家居生态系统会产生大量的实时传感器信息，这些信息通常是相互关联的。这种丰富的相互关联的状态信息成为执行在线分析 ** 和预测的自然驱动力，以改善人群的个性化体验和宏观服务。此外，** 相关状态信息具有固有的部分时间排序，这对于执行时间 ** 分析（如在线异常检测）和对历史信息执行分析（如分析 ** 在一天中的特定时间发生的重复异常）是有用的。Plasmatic的一个重要问题是开发 ** 一个可扩展的系统，该系统将大量相关的传感器状态保持为动态图结构 ** 信息，并通过在线分析特定的时间图程序实现实时分析，同时 ** 还提供对历史图结构信息的批处理。开发这样一个可扩展的 ** 系统是本项目的重点。

Recent advances in Internet of Things (IoT) and smart devices has led to a rapidly growing IoT industry with**global market prediction of $8.9T by 2020 (http://i.sfu.ca/TgFkjR). This industry spans across various domains**from household personal devices to monitoring and management of industry grade large scale equipments.**Plasmatic Technologies Inc. is a Canadian start-up based in Vancouver that enables forward-thinking**organizations and their customers to benefit from Connected Home devices and the valuable data they produce**through Home Insights and Machine Learning. The company's unique Alana Smart Home platform and**applications empower innovative service providers programs that enable differentiation, stronger customer**engagement, revenue growth and operational efficiency. The Alana Smart Home platform and applications are**designed to allow customers to seamlessly interconnect many different smart home technologies to give them**insights and control, while allowing them the flexibility to use the devices they want in their home. Smart**home systems and connected home ecosystems generate a large volume of real-time sensor information that is**often correlated. Such rich interrelated state information becomes a natural driver to perform online analytics**and predictions to improve personalized experience and macro-level services across populations. Furthermore,**the correlated state information has an inherent partial temporal ordering that is useful to perform temporal**analytics like online anomaly detection and performing analysis over historical information like analyzing**repeating anomalies occurring at a particular time of the day. An important problem for Plasmatic is to develop**a scalable system that maintains this large amount of correlated sensor states as dynamic graph-structured**information, and that enables real-time analytics via online analysis-specific temporal graph programs while**also providing batch processing over historical graph-structured information. Developing such a scalable**system is the focus of this project.