CANADA'S LEADING INFORMATION SOURCE FOR THE METALWORKING INDUSTRY

LATEST MAGAZINE

CANADA'S LEADING INFORMATION SOURCE FOR THE METALWORKING INDUSTRY

CANADA'S LEADING INFORMATION SOURCE FOR THE METALWORKING INDUSTRY

How generative AI can transform manufacturing

Share This Post

Many manufacturers already leverage artificial intelligence (AI) across an array of applications to improve their competitive advantage and bottom line by achieving benefits like enhanced forecasting, reduced downtime and increased production precision. The experts at Dell Technologies believe Generative AI (GenAI) can help manufacturers accelerate innovation and make their case in their newreport Generative AI for Smart Manufacturing. Included below are highlights from that report.

While GenAI has enormous potential to transform manufacturing processes and operations, traditional AI has historically supported the majority of manufacturing use cases. These AI models and systems can learn to identify patterns in a manufacturer’s data and use the learnings to make predictions or decisions.

 In recent years, increasing accessibility has greatly improved the appeal of GenAI — including deep learning. Advanced technologies that were once complex and expensive have become more mainstream and affordable, enabling more manufacturers to leverage GenAI systems in innovative ways. The latest GenAI solutions enable manufacturers to utilize their own data and intellectual property to drive improvements in a highly scalable way — one that allows them to maintain their competitive edge and protect sensitive information.

Todd Edmunds, global CTO for manufacturing at Dell Technologies, indicates, “Rapid advances in technology for sustainability, digital twins, and generative AI are converging to create unprecedented opportunities that are redefining what manufacturing can be.”

What makes GenAI different?

Since introducing sensors to the factory floor, plants have collected data to better understand machine performance, product quality, equipment health and energy consumption. Unfortunately, in many cases, that data is in siloes, so it is not used to its full potential, and it may not be used at all.

Generative AI presents the opportunity to utilize extensive data from the entire manufacturing ecosystem to produce net-new outputs, uncovering levels of efficiency and driving extraordinary outcomes that were once beyond imagination.

Rita Wouhaybi is a senior AI principal engineer at Intel Corporation, which partners with Dell Technologies to provide AI solutions that meet customers wherever they may be on their AI journey. She explains, “We have become community data hoarders. We love to collect data. A big opportunity with generative AI will be making sense out of these big chunks of data sitting in private repositories.”

Currently, AI can analyze vast amounts of data with incredible speed. It is often used to define rules that generate descriptive and diagnostic results, such as a production error or a security risk. GenAI allows manufacturers to use data much more ambitiously, enabling a more predictive approach especially when that data is collected with specific use cases in mind.

Wouhaybi adds, “A lot of intelligence can be created by monitoring log files. But IT professionals lose sleep looking at them — like searching for a needle in a haystack. This challenge makes ‘log files’ prime candidates for the summarization ability of large language models. And with GenAI, we can take it further by predicting what is about to happen.”

Many manufacturers have incorporated traditional AI into operational processes and production by integrating rule-based systems and machine learning models that are trained to perform specific tasks or optimize predefined processes. These systems and models rely on explicit instructions and historical data to make decisions.

GenAI takes the power of data further, enabling manufacturers to use their data to train systems and models to identify patterns and, when prompted, generate a new, similar pattern or outputs based on the new pattern.

Edmunds emphasizes that it’s important to ensure you’re able to get enough of the right data for any AI use case. “While plant data may be available for specific use cases, you need to have the systems in place to get it, format it, process it and store it for AI to work its magic.”

Clearly defined objectives are critical

We’re just starting to scratch the surface of how GenAI will impact large manufacturers. As the industry moves forward, applications will roll out quickly. It will be tempting to use GenAI in as many areas as possible.

However, clarity around business goals is critical for successful implementation. Organizations can expect a significant return on investment in GenAI, but they need to be crystal clear on the problem they’re trying to solve, the desired outcomes and the metrics they’ll use to measure success. Only then can they transform their data into the format required to support and address their prioritized use cases.

“Organizations don’t fail because they lack use cases, but because they have too many options,” notes Bill Schmarzo, Customer Advocate for Data Management Incubation at Dell Technologies. “Executives must go through a thoughtful, collaborative process, bringing together all the key stakeholders to identify, validate, value and prioritize the use cases they want to go after — and target use cases they can deliver within 9-12 months.”

Share This Post

Recent Articles




Wordpress Social Share Plugin powered by Ultimatelysocial
error

Enjoy this post? Share with your network