Fluctuating market conditions, supply chain constraints, labor shortages, and a fast-paced global industry are forcing manufacturers of all sizes to reevaluate the way they operate. Many manufacturers have begun to adopt technology to maintain a competitive edge and address long-standing business challenges. From automation to digital technologies, industrial iot, and more, businesses can leverage these innovations to ultimately capture data from diverse systems, processes, and people to provide the strategic insights needed to make better decisions.
There is no doubt that these companies have a lot of data to work with. According to a McKinsey study, manufacturing generates 1.9 petabytes or 1,900,000 terabytes of data annually. The problem was that they needed a better way to capture and analyze data and turn it into usable information, and they needed to do it quickly. As a result, many businesses are turning to artificial intelligence (AI) to find opportunities with their data to improve their operations.
Why is AI perfect for data analysis?
From improving manufacturing yields and uptime, to accurately forecasting demand and remotely monitoring machines, and even controlling assets and improving product quality, AI can be leveraged to significantly improve overall efficiency and productivity metrics.
It's not magic, but a complex set of algorithms that analyze large amounts of data, correlate or learn patterns in various variables, and apply that knowledge to current conditions to help predict future states. This is not to say that humans can't perform these tasks, but that Al can do them faster and process more data with greater precision, improving business outcomes.
For example, in any manufacturing environment, there are traditionally several different workgroups and machines all collecting their own data. The information from each device can vary in quality, format, and timing, which can create obstacles and make it difficult to analyze and glean any meaningful insights from the data.
With the help of AI technology, large amounts of data can be processed quickly, enabling companies to quickly and accurately combine operational information, predict outcomes based on alternatives, and enable manufacturers to make agile, informed decisions. This pre-emptive predictive ability is where AI's strength lies, and it can greatly increase product yields.
By identifying the root cause of product quality problems, AI can help reduce product defects and scrap rates and increase manufacturing yields. With detailed information and analysis, manufacturers can address quality control issues before they directly impact the company's bottom line. Let's look at one such example.
Use AI to improve engine quality
A global engine manufacturer produces large diesel engines for generator sets, naval and Marine applications, and military vehicles. After assembly, each engine is subjected to rigorous testing. During testing, even the most experienced operators often fail to detect subtle signs of a problem, leading to catastrophic failures during testing or once the engine is in service. These failures have caused significant losses, delayed shipments, created backlogged testing areas and upstream production, cost the company millions of dollars annually, and negatively impacted on-time deliveries.
The problem is not a lack of data, but how it is used. In fact, the plant had been collecting process data for years, but only used it for follow-up work after a failure occurred. By looking at the data in this reactive way, the team is unable to understand why these failures are occurring or proactively address them. Ultimately, these issues are seen as a cost of doing business until the company considers using AI on existing data to predict critical asset failures before they occur.
The manufacturer started with a pilot program to lay the necessary data foundation for AI to make an impact. Given the need to use historical data, the company first conducted data cleaning and analysis, with the help of AI, reducing 20 billion data points from 100 engines to 6 billion of the most influential data points in 48 hours.
Next, connect multiple model sets by time and model to visualize the data and identify any data gaps. Based on the gap analysis, adjustments were made to extract certain data more frequently, thereby improving the modeling. By using an AI platform, the entire analysis is done in a low-risk environment without any impact on current production.
From this data, manufacturers are able to establish baselines, identify trends and anomalies, and develop plans to put the information into action. In just a few weeks, they produced a report identifying a group of risk engines by serial number. Based on this information, manufacturers suspect that these engines have a higher probability of problems during quality control tests or in the field. By linking test data to actual product failures, the report accurately identified more than 80 percent of engine problems over several years.
It is important to note that this project is an iterative process, as the AI model is constantly learning. In about 45 days, the model was able to predict failures 30 minutes in advance with a zero false positive rate.
Minimize disruption to operations
During the official launch, the Al solution is connected to real-time data generated by the test control system and human machine interface (HMI). This has no effect on normal operation. In fact, the model had been integrated with the company's standard test software, and the operator was not even aware that it had been implemented. They just need to know that now their HMI interface will inform them of any potential looming issues and how to deal with them.
In the first 90 days, the AI application detected 20 real-time events, avoided more than $4.5 million in engine damage, and achieved a 10x return on investment (ROI) for the project.
As this case illustrates, leveraging AI can provide manufacturers with a way to proactively reduce quality defects, save money, and improve delivery rates while minimizing disruption to operations. Starting with a solid foundation of data and working with experienced partners, AI can provide the insights needed to drive business outcomes and help manufacturers compete in today's rapidly evolving business environment.
But AI doesn't have to be a one-size-fits-all solution. Depending on your needs, application and specific situation, different solutions need to be customized. Therefore, it is important to have a trusted partner at your side. When it comes to AI, they can assess where you are on your digital transformation journey, understand your goals or challenges, and identify the solution from top vendors that best suits your actual needs.
