Service-based Abstraction and Programming Mechanisms for Bridging Business Processes and IoT Big Data

物联网（IoT），即物体、传感器和日常用品的互联，预示着一个人、真实物体和虚拟物体自然融合，并能以方便高效的方式相互交互的新时代。近年来，物联网发展势头迅猛，正在成为各个应用领域的新基础设施。业务流程建模（BPM）已经成为协作流程集成、操作和设计方面的一项强大技术。随着物联网的蓬勃发展，物联网和BPM的协同需求越来越大：BPM可以通过物联网更快、更准确地感知物理世界，快速扩展的物联网大数据可以通过BPM更有效地利用和“编程”。BPM已经广泛应用于零售、物流、制造业和许多其他服务行业几十年了。物联网和BPM是不同的领域，各自开发了不同的概念和形式，但通过形式和方法的融合可以极大地受益于彼此，实际上必须合并才能实现未来的应用。BPM可以为更好地利用物联网数据提供业务逻辑和编程范例。反过来，物联网可以为BPM提供更快、更准确地感知物理对象和环境状态的能力。然而，如何将物联网数据和服务集成到BPM中以使后者能够感知物联网，这是一个挑战。物联网环境结构不良且高度动态。非结构化是物联网的主要特征，因为松散耦合对象之间的大多数通信是临时和情境完成的。使用单块流程模型的传统BPM方法在处理这种情况时遇到了麻烦：编排和重新编排是BPM用于组织和协调业务流的两种主要机制，这两种机制都需要了解参与流程的结构和交互。动态是物联网的常规问题。物联网环境中可能有数以万计的对象，许多对象都是短暂的：一些对象可能会周期性地打开和关闭，新对象可能随时加入，一些对象是可移动的，从一个社区移动到另一个社区，还有一些对象由于系统错误或电池耗尽而消失。在这个项目中，我们的目标是开发必要的编程机制和抽象层，以弥合业务流程建模和物联网应用之间的差距。在我们方法的核心，我们将业务流程模型拆分为几个代表动态物联网组件的微流程。此外，我们将开发转换算法，将物联网业务流程模型转换为可执行实体。

Internet of Things (IoT), the inter-connectivity of objects, sensors, and everyday items, promises a new era where people, real objects and virtual objects are naturally integrated and can interact with each other in a convenient and efficient way. In recent years, IoT has gained huge momentum and is becoming the new infrastructure in various application domains. Business Process Modelling (BPM) has established itself as a powerful technology in the integration, operation, and design of cooperative processes. With the boom-ing of IoT, there is an increasing need to synergize IoT and BPM: BPM may sense the physical world faster and more accurately by means of IoT, and the fast-expanding IoT big data may be more effectively leveraged and ‘programmed’ by means of BPM. BPM has been widely used in retailing, logistics, manufacturing and many other service indus-tries for decades. IoT and BPM are different areas with separately developed concepts and formalisms, but could greatly benefit from each other through a convergence of formalisms and methods, and in fact have to be merged to enable future applications. BPM can provide business logic and program-ming paradigms for better utilizing IoT data. In turn, IoT can equip BPM with the ability to sense the status of the physical objects and environments faster and more accurately. How to integrate IoT data and services into BPM to make the latter IoT-aware, however, is challenging. The IoT environment is poorly structured and highly dynamic. Unstructuredness is the principal characteristic of the IoT as most of the communication between loosely-coupled objects is accomplished ad-hoc and situative. Traditional BPM approaches with monolithic process models have trouble dealing with such situation: both orchestration and cho-reography, which are the two main mechanisms of BPM for organizing and coordinating busi-ness flows, require knowledge about the structure and interactions of participating processes. Dynamism is a matter of routine for IoT. There might be tens of thousands of objects in an IoT environment and many objects are transient: some objects may turn on and off periodically, new objects may join at any time, some objects are mobile and move from one community to the other, and some objects just disappear due to system errors or running out of battery. In this project, we aim to develop the necessary programming mechanisms and layers of abstraction to bridge the gap between business process modelling and IoT applications. At the core of our approach, we split business process models into several micro processes that represent the dynamic IoT components. We furthermore will develop transformation algorithms to convert IoT business process models into executable entities.