Seeing Beyond the Dashboard: Insights on Industrial AI

Many of the challenges organizations face when adopting AI echo earlier industrial revolutions. What matters now is what we’ve learned from them. A couple of weeks ago, my colleague Diane Abbott, associate director of executive programs, held a thought-provoking conversation on LinkedIn Live with Senior Lecturer John Carrier on the topic of Industrial AI. Listening to John, it was easy to see how the latest advances in AI connect to the time-tested principles of systems thinking and control theory, yet his linkages may not be as apparent in industrial operations or boardroom meetings.

Here are five takeaways that stood out to me from John’s remarks.

1) Dashboards don’t replace the factory floor.

It may be tempting to believe that a spreadsheet or a beautifully designed dashboard tells us everything we need to know. Yet as John put it, “fantasy football is not football.” Metrics and dashboards must reflect the realities of the shop floor, the service line, or the operating room, not only the financial indicators that executives prefer to track. Leaders should ask: Are our dashboards providing the right information to the right people at the right time?

2) Data without focus can be dangerous.

We often hear that “data is the new oil.” But like oil, data in the wrong place can wreak havoc. John shared the sobering example of a refinery incident in which 4,000 alarms were triggered before a fatal accident, so many that the critical signals were lost in the noise.

More data does not automatically mean more insight. Organizations must prioritize leading indicators—signals that help anticipate and prevent problems—even if those indicators reveal uncomfortable truths. Building a culture that accepts and acts on “bad news” is just as important as collecting the data in the first place.

3) Think in terms of control, not just AI.

Industrial AI is best understood not as a collection of algorithms, but as a way to control complex systems better. The goal is not to “buy AI,” but to improve profitability and reduce risk by making operations more controllable. As John noted, even some of today’s most advanced AI approaches, like generative models, can be understood through decades-old concepts in control theory, reminding us that proven principles still matter.

4) Faster time constants bring a competitive advantage.

One of John's most memorable points was about “time constants,” which refer to the actual pace at which processes occur in a system. Too often, organizations assume they know how long things take, only to find that their estimates are wildly optimistic.

This matters because customers demand precision and reliability. The organizations that win are those that can observe, orient, decide, and act faster than their competitors. AI can play a vital role here, especially in observation and orientation. But the underlying objective is always the same: shortening the cycle time to respond more quickly and accurately.

5) Keep it simple and reward results.

Finally, John emphasized the importance of simplicity. Complexity in robotics, language models, or process automation often leads to fragile systems that fail under real-world conditions. Sometimes the most effective solutions are not glamorous. The real challenge for leaders is to celebrate these “boring” wins, rather than rushing on to the next shiny technology.

Looking Ahead

As John concluded, the winners in the age of industrial AI will not be the companies that purchase the most advanced tools, but those that integrate them most effectively into their culture, leadership, and operations. It is about aligning technology with human creativity, shortening decision cycles, and focusing on leverage points where small changes can yield outsized results.

Ready to dive deeper? John teaches an executive education course, Strategy, Survival, and Success in the Age of Industrial AI. Visit executive.mit.edu/ind to learn more.