GOALI: Modeling and Control for Manufacturing Intelligence with Cloud Computing and Storage

这个学术与工业联络（GOALI）项目的资助机会将使用基于云的框架来研究制造系统操作的范式转变，以实现更高的质量，更高的吞吐量和更低的能源使用。一个新兴的行业趋势是将制造工厂车间产生的大量数据存储在云中。这些数据包括生产目标、质量检测结果、机器故障等，但也可能包括车间每台机器的详细能源使用信息以及每台机器的维护日志。 在目前的实践中，这些数据被收集到仪表板中，工程师可以使用这些仪表板来了解制造系统的当前状态。 在这个项目中，我们将使用这些数据来自动查找质量、吞吐量、维护和能源使用之间的关系。 可以分析和利用这些关系来制定更好的生产计划，从而降低整体生产成本。为了实现这一目标，将开发制造系统中不同组件的模块化混合模型。 这些混合模型将包括组件的离散事件行为以及它们的连续变量动态。 利用工厂车间产生的运营数据并将其推送到云存储中，将开发自动方法来从实时数据流中提取这些模型的参数。最后，将开发优化功能，可以编码生产率，能源使用和维护之间的权衡，开发的模型将用于解决最佳生产参数。 这项研究将改变制造运营的艺术状态，实现高度的维护调度和定制，以满足制造运营商不断增长的需求。所产生的关于制造系统的混合建模和多目标系统优化的知识将通过云的有效和高效的监测和控制，导致智能制造的新范式。此外，这项研究将为完全自动化的基于云的制造操作奠定基础。

This Grant Opportunity for Academic Liaison with Industry (GOALI) project will investigate using a paradigm shift in manufacturing systems operation using a cloud-based framework to achieve higher quality, higher throughput, and lower energy usage. An emerging industry trend is to store the large amount of data produced on manufacturing plant floors in the cloud. This data includes production targets, quality inspection results, machine faults, etc, but has the potential to also include detailed energy usage information of each machine on the plant floor as well as maintenance logs of every machine. In current practice, this data is collected into dashboards that engineers can use to understand the current state of the manufacturing system. Within this project, we will use this data to automatically find relationships that exist between quality, throughput, maintenance, and energy usage. These relationships can be analyzed and leveraged to develop better production schedules, reducing the overall cost of production. To accomplish this goal, modular hybrid models of the different components in the manufacturing system will be developed. These hybrid models will include both the discrete-event behavior of the components as well as their continuous-variable dynamics. Leveraging the operational data produced on plant floors and pushed into cloud storage, automatic methods will be developed to extract the parameters for these models from real-time data streams. Finally, optimization functions will be developed that can encode the tradeoffs between productivity, energy usage, and maintenance, and the developed models will be used to solve for the optimal production parameters. This research will transform the state of the art of manufacturing operations, enabling a high degree of maintenance scheduling and customization to meet the ever-increasing demands of manufacturing operators. The generated knowledge on hybrid modeling and multi-objective system optimization for manufacturing systems will lead to a new paradigm of intelligent manufacturing through effective and efficient monitoring and control with the cloud. Furthermore, this research will lay the foundation towards a completely automated cloud-based manufacturing operation.