There are many categories of Digital Twins, each with their specific value. These include component, asset, system, process, or an entire manufacturing plant. The Digital Twin of each of these categories can include data from other categories; for instance, the Digital Twin of an Asset includes data from Components, and can include assembly information from System.
- Component/Part: The basic building blocks of electronics: transistors, resistors, capacitors, regulators, etc. These can be serialized or (more typically) non-serialized
- Asset: A serialized unit which performs a discrete function in and is an integral piece of a higher level assembly. An example would be a printed circuit board (PCB).
- System: A serialized unit that is comprised of multiple constituent assets. There can be many system levels and genealogy (for instance, a higher level assembly can be made of several assemblies, each of which is made of several assets). An example can be an aircraft engine control system.
- Process: The steps and procedures taken during a unit’s lifecycle, each of which comprises one or multiple events. Manufacturing is one example, MRO is another.
- Manufacturing Plant: The location where manufacturing processes are located. These include test, assembly, rework and other related processes.
In discrete manufacturing of complex electronics, a system digital twin is the most useful approach to optimize manufacturing efficiency, throughput, and product quality. The major challenge of developing a system digital twin is the convergence of information across many complex enterprise information sources, including product design, ERP, test, MES, supplier quality, and service into a single source of truth.
Principal Industry Lead-Azure
Microsoft
Considering Diego’s definition, what is the relevance of the digital twin to electronics manufacturing, and how can we apply it?
To put it succinctly, it means having the data of the as-built and as-maintained object from all areas where data can be gathered, and (perhaps most importantly), making the data analyzable and useful to drive better business outcomes.
At a corporate level, executive decisions on cycle time and utilization rates help drive overall profitability and success. By linking data from each manufacturing stage and event in the same normalized data warehouse, users can have two key deliverables: how a metric (like cycle time or utilization rates) is performing and trending, and the underlying reasons why that metric is performing the way it is. The latter deliverable gives technical users and engineers the data and tools they need to investigate, resolve and/or prevent the former deliverable.
These deliverables (executive decisions on KPIs with root-cause workflows) rely on a system that can provide a foundation for the Digital Twin. IntraStage BlackBelt Enterprise leverages a web enabled architecture based on Microsoft Azure technologies, as well as a purpose built manufacturing data model that ultimately enables distributed data democratization.
Let’s examine a few examples of IntraStage customers using that Digital Twin foundation and data democratization to identify and resolve some key performance indicators.
Bold corporate-level initiatives like driving towards zero PPM or DPMO require aggregation of data from various sources. This often results in an implementation of a common Digital Twin data model that scales across product lines, manufacturing locations, and supply chains.
Supplier quality isn’t measured solely from supplier scorecards, incoming inspection, or inspection at the source. The quality of supplier parts needs to be gauged in manufacturing to see what is failing via PPM and DPMO. These metrics drive product quality and operational efficiency.
Overall Equipment Effectiveness (OEE) and Cycle Time are critical metrics describing how a product is performing in its manufacturing process, and the efficiency and speed of that process and its constituent machinery. Executive leadership needs these metrics to maintain profitability, efficiency, and manufacturing velocity.
With BlackBelt Enterprise, users can have real-time insight into analytics that drive these metrics and have root-cause workflows to identify, investigate, and resolve any OEE, Cycle Time, and utilization issues.
Having high-level metrics is good; having high level metrics integrated with workflows that leverage the underlying details that drive and provide insight into the metrics is better.
A board overlay is a prime example of having multiple sources of data inform a workflow that outputs in better quality and processes.
For complex electronics manufacturing, there are several common terms for the configuration of a unit: as designed, as built, as maintained
Another term we use at IntraStage is ‘as-is’.
