Towards a Contextualized Visual Analysis of Heterogeneous Manufacturing Data
Towards a Contextualized Visual Analysis of Heterogeneous Manufacturing Data
复制标题
对异构制造数据进行情境化可视化分析
DOI：
10.1007/978-3-642-41939-3_8
复制
发表时间：
2013
影响因子：
3.2
通讯作者：
B. Urban
中科院分区：
文献类型：
--
作者：
Mario Aehnelt;Hans;B. Urban
跨多个数据源的可视化分析通常需要集成多个专门的应用程序来处理它们的异构性。在制造业中也是如此，必须对订单、人员、工作负载、维护等数据进行综合分析，以做出有根据的管理决策。然而，多个数据源和应用程序的编排给软件基础设施和分析师带来了挑战。我们提出了三个层次的方法来科普这些挑战。在第一步中，我们建立了一个领域相关的工作流作为分析师的心理模型。基于上下文化的新概念，我们将不同的应用程序与此模型进行了有意义的整合。在第三步中，我们通过基于服务的体系结构，根据数据在对齐的应用程序中的使用来合并数据。通过在用户级别开始集成，我们能够务实地确定目标，并将所需的集成简化到技术上可实现和交互式管理的程度。我们使用Plant@Hand系统来整合制造数据和应用程序。
Visual analysis spanning multiple data sources usually requires the integration of multiple specialized applications to handle their heterogeneity. This is also true in manufacturing, where data about orders, personnel, workloads, maintenance, etc. must be analyzed together to make well-founded management decisions. Yet, the orchestration of multiple data sources and applications poses challenges to the software infrastructure and to the analyst. We present a three-tiered approach to cope with these challenges. In a first step, we establish a domain-dependent workflow as the mental model of the analyst. Based on the novel concept of contextualization, we then align the different applications with this model for their meaningful integration. In a third step, we incorporate the data according to its use in the aligned applications by means of a service-based architecture. By starting the integration on the user level, we are able to pragmatically target and streamline the required integration to a degree that is technically achievable and interactively manageable. We exemplify our approach with the Plant@Hand system for integrating manufacturing data and applications.