Get to know Bob Latino:

Profile: linkedin.com/in/boblatino

Website: prelical.com

Email: [email protected]

In his conversation on the Counter Errorism podcast, Bob Latino shares valuable insights into defect elimination, system reliability, and the evolving role of technology in incident investigations.

Here are the key lessons we learn from his approach:

Experiential Learning and "The Manufacturing Game"

Latino advocates for using experiential learning to teach defect elimination, specifically highlighting a simulation called the "Manufacturing Game" originally developed at DuPont. While leaders are often initially skeptical about paying their employees to "play a board game," the exercise effectively models the complex system dynamics of a real plant over a day and a half to two days. The simulation throws real-world variables at the players—such as safety violations, defective parts, and customer complaints—and proves that an organization will only survive if its operations, maintenance, and support services learn how to effectively communicate with one another. Because workers actually experience these lessons, the principles become highly memorable and impactful.

The Link Between Reliability and Process Safety

Latino highlights a strong, historically proven correlation between equipment reliability and process safety. He explains that when a plant achieves "steady state operation" through good preventive and predictive maintenance, it experiences far fewer unexpected upsets. Consequently, workers are much less likely to be forced into dangerous, energized areas to address sudden mechanical failures, which inherently keeps them safer.

The Limits of AI in Root Cause Analysis (RCA)

When looking at the future of incident investigations, Latino warns about the severe blind spots of Artificial Intelligence. While AI is excellent at analyzing the mechanical "physics of failure" (such as detailing the various ways a specific bearing might break), it completely lacks the critical real-world context of an event. AI cannot comprehend the nuances of complex human decision-making or flawed organizational systems—such as a worker making a mistake because they were forced to use an obsolete procedure that management failed to update after installing new technology. Latino advises that investigation teams should exhaust their own human knowledge and understanding of the context first before turning to AI, rather than using it merely as a shortcut to rush a report by the end of a shift.

Mentions in this episode:

The Manufacturing Game

Please check out the amazing people doing amazing things on these podcasts we listen to and love: