Modeling and simulation: Achieve next-level results with AI
Aerospace executives can now optimize manufacturing processes by leveraging artificial intelligence (AI) with high-performance computing (HPC) technologies and the digital thread. A digital thread follows the lifecycle of a product from design inception through engineering and product lifecycle management, to manufacturing instructions, supply chain management, and through to service events. You'll be able to enhance the aerospace design process to protect budgets, avoid static production rates, and nudge your business ahead of competitors.
Even better, as aerospace design becomes more complex, AI can help keep your business ahead of the innovation curve.
Here are three ways to get the most from using model-based systems engineering (MBSE):
1. Realize bigger wins from the start by aligning stakeholders
I recently chatted with a vice president of IT Infrastructure at a large manufacturing company and his message was very clear: HPC technologies that support modeling and simulation are very important to his business users. He explained that even though engineering teams tend to fully consume (and often exceed) available resources, the challenge is ensuring effective use of those resources.
Modeling and simulation benefits typically include:
- Accelerated innovation by digitally depicting physical and logical systems
- Improved quality by analyzing data generated during design and production
- Reduced costs by demonstrating failures in product designs so you don't build them
- Accelerated delivery due to the ability to test multiple scenarios in hours vs. days
A lot of IT investment goes toward modeling and simulating new designs with the goal of delivering products that are faster and cheaper.
However, while modeling and simulation activities do drive consumption of expensive resources and much work goes into proving out new designs, cost benefits aren't always clear.
This happens mainly because what is proved out in modeling and simulation can't be delivered due to differences in the motivations of stakeholders.
While the investment, business, and technology communities are aligned on the vision, each silo has a different focus.
Manufacturing executives drive risk and investment decision-making, so they see the costs and benefits of a resource in such terms. IT provides the technology, but it doesn't directly generate or receive the financial benefit of what HPC or AI resources provide to the business. Engineers innovate using IT capabilities, but they are not as concerned about the costs to provide resources. They are more focused on providing innovation or improvement in product design and delivery.
2. Drive unity in the design process by incorporating a digital thread
Historically, aerospace manufacturers leveraged analytics and data collection to improve industrial operations and drive innovation. Yet, one aerospace director recently said, "The factory of the future continues to generate lots of precious information via connected objects that doesn't contribute to the digital thread. This data can be analyzed to understand the root cause of product deviations, to fight scrap rates and tackle quality issues, while making sure the machines are utilized at full potential, and to increase productivity."
However, a digital thread that depicts a data flow and integrated view of data throughout each value stream can help. A digital thread is traceable across the lifecycle of a product, such as in the design, performance data, product data, supply chain data, and software extending into the product created.
This is where connecting IoT with operational technologies (OT) is a game changer.
OT includes those hardware and software components that monitor and control physical devices, usually connected to compose a digital thread.
A simple example of OT connected as IoT is where a 30-year-old fluid pump that's integral in the creation of parts and critical to the success of the manufacturing process holds valuable operational information yet doesn't get shared with up- and downstream disparate processes. Sharing that data could offer insights to improve such a design.
Sometimes the amount of investment in an IT resource that collects data for a single OT function may be more than the value of the OT process in which it participates. However, collecting data from many OT functions and contributing that to a fluid data stream, or a digital thread, offers a huge opportunity to realize massive benefits from modeling and simulation. These could include increased performance of the part while operating and optimized maintenance tasks to reduce downtime while in production.
Aerospace manufacturers that use a digital thread concept with an MBSE strategy will be able to control data analysis throughout their product design lifecycle so it is later processed into valuable information. This is the solution to the stakeholder differences mentioned earlier.
This is also where understanding approaches to AI can bring even greater returns.
3. Strategically manage data to enable AI
Data collection can be very difficult, requiring an investment and technology strategy that can be adopted quickly and effectively without impacting the manufacturing rate of a business.
Take a common-sense approach around data when building AI into your modeling and simulation platform while maintaining a digital thread that contributes to the business.
Find the data
There is an endless amount of data available. Identifying what data you need and using it at the right time can be very challenging. Leveraging a well-defined data services architecture will accelerate building models that can contribute to your AI strategy.
To prepare for AI, evaluate your design process and data sources to identify functions that could benefit from analytics, driving an immediate result with minimal change. Determine if you have the data you need and how you will integrate the benefit back into your design.
Collect the data
Placing IoT data collection and inference technology closer to the manufacturing process enables near- and real-time inference vs. the traditional capture-and-store historian model. Establishing a technology framework is a key building block for ensuring data gets to the right place at the right time.
Integrate the data
Assuming you get past data collection via IoT/OT approaches, you now must integrate and manage the data and make it usable for the consumers up- and downstream. Ultimately, you want to build a digital twin of each process and then a macro view of how the discreet processes are dependent or interdependent.
Last, prioritize the results to create a self-funding strategy to expand beyond the initial focal area. This all leads to connecting all the stakeholders in a strategy they understand and will support.
Incorporating the digital thread concept
Although we live in a digital world today, many in manufacturing are hamstrung to execute on what appears to be an invaluable transition to a completely digital model and simulation platform. Most businesses have discreet organizational and financial strategies to mitigate this, which is normally a three- to five-year roadmap. However, people and technology change so fast that the roadmap frequently never gets executed.
In aerospace manufacturing, achieving the accelerated results depends on reducing the number of physical tests and replacing them with virtual tests. The digital thread concept makes this possible by sharing and ultimately cutting down development costs and lead times. Therefore, incorporating the digital thread concept so that data can be used to connect disparate components and interdependent processes enables next-generation simulation and modeling results throughout the business.
AI approaches can bring cost, production, and competitive benefits to the MBSE processes.
The digital thread in the aerospace industry
This article/content was written by the individual writer identified and does not necessarily reflect the view of Hewlett Packard Enterprise Company.