Digital overload vs. value creation

With the massive influx of nerd-like terminologies – digital this, AI that – it’s natural that potential beneficiaries in manufacturing have become overwhelmed, especially when seeing very few of their pioneering peers achieve real success. Such risk is not acceptable in the medical device industry, so how do we make digitalization beneficial for product reliability?

Data may be used in many ways within the organization to achieve different things. If subsequent genuine accumulated value is achieved, the cost of data is very low. Let’s start at the top of our data value-chain.

Digitalization starts with design. Proving product designs using samples and trials is costly, leading to extended lead times and limited scalability for designs with many variants or are custom in nature. Moving the visibility of the designed product into the digital realm removes the need to have anything physical for confirmation purposes and yet provides the level of confidence needed to go forward into production. While designers get a significant benefit from the reduced lead time and effort per product, another result is a dataset for each design, which is effectively free for use in manufacturing.

The use of CAD data in manufacturing differs significantly from design. For manufacturing, CAD data represents a full and detailed breakdown of the product to be made, describing the full expectation of what it is and how it is to be produced. Merging the CAD data with the bill of materials (BOM), digital manufacturing engineering can compose the necessary production tasks to be done in sequence and assign it to a set of production operations, automated and manual.

Without digitalization, this is very much a manual effort, dependent on retained skills of engineers to understand the nuances of product requirements against production station capabilities, especially where there are complex dependencies and sub-assembly hierarchies. In the digital world, this knowledge is captured within the engineering tool. Multiple-use operational modules and templates are prepared based on a standard set, then simply configured by engineers. Operations are then composed by combining appropriate operations to which product data will be automatically applied, according to the specific product and production configuration.

Being automated in this way preserves the know-how with which assignments are made but takes just a few seconds rather than many hours, as well as avoiding potential mistakes or issues often overlooked, which are natural to an intense task performed manually. Further benefit is achieved as engineers are freed from repetitive tasks, to contribute further to the refinement of engineering rules and practices, increasing product quality and reliability, as well as decreasing lead time, scrap, and other costs. The operational production data created is now effectively available free for use in manufacturing.

Deploying data collection technology

The use of engineering data in manufacturing essentially provides the context against which data collected from manufacturing stations is constructed. Such data collection is best done using modern Industrial Internet of Things (IIoT) standards, such as the IPC Connected Factory Exchange (CFX), which is plug and play. The best manufacturing execution system (MES) solutions also provide options of hundreds of custom interfaces and support for older standards as part of their solution.

The important aspect is the ability to rapidly deploy data collection technology without depending on third-party middleware or excessive customization which, if unchecked, can result in unexpected costs and compromise. The aim is to automatically process the transactional data from manual and automated stations accurately in context with the product and BOM data, such that automated analyses and decisions can be taken.

Simple examples of the hundreds of such values include verification of each product along its prescribed path, the Just in Time (JIT) preparation and verification of materials at each production station, the active assessment of quality trends, and the ability to optimize operations as they’re working. Such automation removes the risks associated with human operations. The resulting assurance contributes strongly to quality and reliability through elimination of mistakes and proven adherence to procedures. In addition, contextualized material and process data is now freely available for advanced analysis.

Finding, avoiding defects

Market quality and reliability have been proven to be directly proportional to manufacturing quality. Even a single defect discovered during the manufacturing process, which could be relatively simple and inexpensive to fix, may represent the tip of the iceberg of underlying product weakness across products that didn’t quite fail in manufacturing but were to become statistics of unreliability in the market. Tracking down the root-causes of defects is essential.

Using the contextualized traceability data, analysis can reveal potential defect causes based on complex patterns of results. A unique set of conditions from multiple processes can contribute to the creation of a single defect. It’s easy then to discover through analysis products which were exposed to very closely the same conditions, and so would be likely candidates for market failure. Intelligent, guided inspection within the manufacturing operation is orders of magnitude less expensive than dealing with market failures and potential consequences. With automation, these conditions may also be detected and avoided in future manufacturing.

All of this could theoretically be done by a team of experienced manufacturing professionals, skilled and knowledgeable in industrial engineering practices. However, this isn’t financially viable given the sheer volume, complexity, and variation in today’s manufacturing unless a very large part of work is automated using digitalization. The orchestration of all of this within manufacturing is an advanced IIoT-based MES solution that’s inclusive of all major elements that directly influence the efficiency of the operation and quality of the product. Any discrete manufacturing process can be elevated through using this data and intelligence.

Adopting digitalization

What’s described is a cost-free introduction and adoption of digitalization. At each stage, data has been accumulated so it benefits each part of the operation and is available for the next, all while being contextualized within a single data model. With any professional software there are licensing and deployment costs, but these become trivial when digitalization’s approached correctly, avoiding the unexpected burden of middleware and customization. Values and benefits far outweigh initial costs, creating a compelling return on investment (ROI).

Most vendors involved in digitalization focus on technology and assume customers will figure out the values themselves. They sell individual roads and highways without any consideration for vehicle types, quantities, traffic patterns, or navigation, which represent actual value.

With a success rate across the industry as little as 15%, medical device manufacturing isn’t the arena in which new technologies – digital this, AI that – should be trialed. Instead, such technologies should be introduced based on the proven needs and manufacturing flow, respecting the life of the products and their customers. During your important and inevitable digital transformation, don’t follow the buzzwords blindly with solutions hoping to be a step toward some benefit. Stick to your principles and let the value do the talking.

About the author: Michael Ford is senior director of emerging industry strategy at Aegis Software.

Aegis Software
https://www.aiscorp.com