Product Lifecycle Data Exploration and Visualization

Summary

To develop and deploy advances in methods, standards, and software tools for data visualization and exploration for improving the efficiency and agility of smart manufacturing systems.

Description

With the continuing emergence of low-cost, lightweight, and powerful hardware at the edge, opportunities grow for advancing methods, standards, and software tools for data exploration and visualization in smart manufacturing systems.  Small and medium-sized enterprises have limited resources to effectively implement standard and reproducible analyses of their manufacturing operations to draw timely and actionable insight. Without formal standards, tacit knowledge embedded in the most experienced employees is not easily transferrable to novices.  Given the recent advancements in computing efficiency (hardware) and intuitive visual interfaces (software), techniques for interactive data exploration can help close this gap.  The goal of this project is to develop and deploy advances in methods, standards, and software tools for data visualization and exploration for improving the efficiency and agility of smart manufacturing systems.  To support this goal, the project will focus on developing, curating and disseminating reproducible, reference implementations of manufacturing data pipelines for improving the efficiency and agility of manufacturing operations. 

WHAT IS THE TECHNICAL IDEA?
Visual analytics, defined as “the science of analytical reasoning facilitated by interactive visual interfaces,” facilitates the decision-making loop by augmenting the data exploration process.  Visual analytics is particularly appropriate in the context of large and complex data sets.   The manufacturing sector generates over 2 Exabytes of data a year (McKinsey Report 2010).  Visual analytics-based approaches for interactive data exploration in the manufacturing domain will (1) enable more informed decision making for operational control, (2) facilitate communication within and across engineering teams (e.g., design, manufacturing, sustainment), (3) aid the exchange of manufacturing data and analysis from one stakeholder to another, (4) equip less experienced newcomers with augmented views of the manufacturing systems, and (5) augment various viewpoints across the lifecycle (e.g. relating procurement capabilities to product specifications).

To exploit the core concepts of visual analytics for managing and understanding product lifecycle data, this project will focus on creating methods, guidelines, and toolkits to quicken the exchangeability and uptake of visualization software components.   The use cases will focus on three research thrusts including monitoring manufacturing activities, mapping disparate lifecycle model definitions, and exploring system tradespaces.  

WHAT IS THE RESEARCH PLAN?
Product lifecycle data is vast, uncertain, complex, multi-modal and sourced from heterogenous data sources.  Visual analytics has shown significant promise for facilitating stakeholder reasoning in environments with similar characteristics.  However, there is a need for additional research and guidance for how best to leverage such methods in smart manufacturing systems.  In response, th research will focus on three core thrust areas: 

1. Smart Manufacturing System (SMS) Monitoring through Interactive Visualizations: (a) open-source software to support monitoring devices. (b) interactive visualizations to efficiently view standards-based machine data in near real-time. (c) standards requirements for the use of augmented reality and virtual reality to support SMS.
2. Mapping across Standard Product Definitions: (a) software supporting common basis for comparing standard product definitions across the lifecycle (e.g., STEP, QIF, MTConnect, ASTM E3012-16, Ecospold). (b) interactive visualizations for exploring relationships across such representations. (c) recommendations for standards improvement or development to enable scalable mapping. 
3. System Tradespace Exploration: (a) inferring similarity of manufacturing process capabilities based on historical product lifecycle data. (b) exploration of experimental data for developing domain-specific (or facility-specific) manufacturing-oriented strategies. (c) visualizations to facilitate the exploration process across multiple domains, e.g. product development activities, manufacturing execution environments, and supply chain decision making.