Ready to Harness the Power of AI in Root Cause Analysis?

Spend less time manipulating data and more on solving problems

Published: Tuesday, August 22, 2023 - 12:02

Solving problems goes beyond noticing the symptoms and wanting to resolve them. It’s also necessary to perform a root cause analysis, pinpointing the factors likely to have made an issue occur. It’s only then that leaders can create concrete solutions for lasting changes. However, root cause analyses can be extremely time-consuming. Fortunately, artificial intelligence (AI) can provide high-speed data processing capabilities that make these investigations more efficient.

Receiving the knowledge to cope with unwanted circumstances

When modern manufacturers see sales trends going in the wrong direction, they must act fast to determine why. Is the issue due to a short-term matter, or is the consumer disinterest a result of a prolonged problem? 

AI works well for cutting through all the data noise that could lead humans to the wrong conclusions. It can also reduce information-preparation errors that make the findings incorrect or less useful. Consider the case of USAopoly, a game and puzzle manufacturer: The company dug into data to determine root causes, but it took a prohibitively long time.

AI can take the drudgery of data collection and preparation out of the hands of humans, allowing them to do what they do best—solve problems.

“The typical process of working with data involves spending about 80% of your time preparing the data and just 20% of your time doing the actual data analysis,” explains Eric Richardson, manager of forecasting and data analytics at the organization. He discussed how his team used to spend six hours per week gathering point-of-sale data from all retailers and putting it into a forecasting template.

The company now uses an AI tool that handles many of the preparation steps. It allows the team to spend more time analyzing the data rather than getting it ready for examination. They could then devote more work hours to tracking downturns. That change led to numerous improvements, such as a positive outcome for a root cause analysis concerning declining sales.

USAopoly achieved a six-figure sales increase and saved more than $300,000 through efficiency enhancements by using AI. The technology also indicated areas for improvement, allowing decision-makers to act quickly with that information.

Succeeding with using AI for root cause analysis requires people to have specific goals in mind. One instance involved an executive at ARM using artificial intelligence for faster chip design. He said that approach generated significant time savings when finding the root cause of problems.

Reducing equipment failures

Repetitive breakdowns of vital machinery can be extremely costly and disruptive. Many leaders mitigate those instances with predictive maintenance. This approach typically involves smart sensors and AI algorithms. Together, those technologies can warn people of failures before they happen.

Some of the information gathered by the AI algorithms may assist people in performing a root cause analysis, too. For example, they may conduct oil-condition tests on a piece of machinery that has frequently broken down over the last few months. If it shows the oil is dirty or the wrong type for the equipment, those problems are easy to fix.

AI is also useful in encouraging companies to adopt data-driven methods for failure identification. In one case, an aerospace manufacturer had ongoing communication problems with satellites that sent data to Earth through a ground station. These issues resulted in lost customer data, and some satellites could only make one transmission attempt per orbit.

Engineers initially had to manually troubleshoot the matter, which often required checking through field notes and log files. However, diagnosing and rectifying the issue this way took too long, which led to those involved becoming interested in AI.

The eventual solution used artificial intelligence to immediately identify failures and the suspected root causes. That information made it much easier to resolve the problems faster. Engineers also relied on programmatic labeling to categorize the faults and their reasons, which improved overall data quality.

One commercial product from Exalens combines root cause analysis with preemptive measures to improve cost-effectiveness. “We’ve leveraged cutting-edge AI technologies to develop an OT endpoint detection and response solution with a complete picture of network and process endpoint activity that will significantly reduce unplanned downtime, saving manufacturers potentially thousands, if not millions,” says Ryan Heartfield, the company’s CTO.

Improving workplace risk visibility

Statistics indicate there could be as many as 2.1 million available manufacturing roles by 2030. As decision-makers work to overcome the deficit, they must consider the best ways to make their work arrangements appealing. Doing that should increase candidate interest while reducing turnover. Perks like above-average salaries and generous benefits can help with these things, but it’s also necessary to maintain a safe workplace.

Darrel Nickerson is the director of health and safety at Irving Forest Services. He explains how his company gathers hazard identification reports from all employees. The company then uses the data from those documents to help the organization be more proactive in accident prevention.

“We discovered that a significant portion of injuries happen during routine work,” says Nickerson. “This finding has prompted us to dig deeper into the root causes and examine the potential impact of complacency among employees engaged in repetitive tasks. It’s essential to understand these patterns to implement targeted interventions and mitigate risks effectively.”

He’s also upbeat about the potential of AI to enhance current processes. His company doesn’t use it yet, but Nickerson envisions using equipment such as real-time cameras to gather data about environmental risks or safety violations, supporting the current processes.

That’s already happening in some places. At a Singaporean construction site, more than 90 AI-equipped cameras analyze the environment. One health and safety officer there estimated a 70% improvement in identifying unsafe working conditions after deploying the technology.

Using the cameras is also vital because it takes safety personnel 45 minutes to perform a site walk. The massive grounds provide ample opportunity for things to go wrong without a manager noticing. However, AI delivers additional visibility for greatly improved safety. It’s equally applicable to expansive warehouses and manufacturing facilities.

Root cause analysis made better with AI

The real-world examples here give compelling reasons for using AI to determine why things go wrong. That knowledge is essential for enabling continual improvement and company growth. Artificial intelligence can find patterns in massive quantities of data, leaving humans with more time to address the problems and prevent them from happening again.

Before investing in a specific solution, people should always identify which parts of their current process take the most time or often lead to dissatisfaction. That’ll help them use AI most effectively and strategically.

About The Author

Emily Newton

Emily Newton is the editor-in-chief of Revolutionized, an online magazine exploring the innovations disrupting the scientific and industrial sectors.