Ahmet Ö.
Corporate
- Thread Author
- #1
Don't Let Data Abundance Turn into Decision Scarcity!
In today's manufacturing world, thanks to automation, we have more data than ever before. Machines constantly generate data, and systems record operating conditions, maintenance events, quality results, and process deviations. Sensors can detect changes long before problems escalate into a noticeable failure. However, having more information doesn't automatically mean better decisions are made.
If a system repeatedly detects the same problem, but no one changes the process, the organization is collecting information but not learning from it. If a maintenance team finds a recurring failure, but this information doesn't reach engineering or production, the same problem can reappear repeatedly. And if a process changes, but no one tracks what happens afterward, there's no way to know if the change worked.
NIST defines smart manufacturing decision systems using a "data feedback loop" that models, senses, transmits, analyzes, communicates, and acts on data. What I've learned after nearly three decades working with industrial equipment and legacy infrastructure is that the power of this loop is more important than the amount of information an organization collects.
Turning Automated Signals into Decisions 💡
Automated systems can tell a manufacturer that a temperature has changed, a machine has stopped, or a process has gone outside its expected range. But a single data point rarely explains why something happened or what needs to change next.
This requires context from the people who know the equipment:
- A technician might know that an alarm tends to occur after a component has begun to wear.
- An operator might notice that a process variation occurs under specific production conditions.
- A supervisor might see that several quality issues share the same root cause.
Schneider Electric's smart factory in Lexington, Kentucky, offers an example of bringing these resources together. The company connects equipment information, analytics, and operator tools through its EcoStruxure platform. Its Augmented Operator Advisor provides employees with equipment information as they work on a machine. Schneider reported that this technology reduced the average repair time on critical equipment by 20%.
The company also used AVEVA Insight to bring together operational data from disparate systems, reducing downtime in critical processes by 5% and achieving a return on investment in less than six months.
In this use case, technology accelerates the path from detection to investigation to response.
What Does a Closed Feedback Loop Look Like in Practice? 🔄
Maintenance illustrates the difference between collecting data and learning from it. If a component fails, replacing it gets production back up and running. If the same component fails repeatedly, the manufacturer needs to understand what happened before the failure, whether warning signs could have been detected earlier, and whether a change would reduce future failures.
Sachsenmilch, one of Germany's largest dairy producers, combined information from existing control systems with vibration monitoring sensors and Siemens Senseye Predictive Maintenance software. The system detected that a pump was nearing the end of its service life, giving the company time to replace it during planned maintenance rather than waiting for it to fail during production.
Sachsenmilch officials state that this single intervention prevented a longer shutdown and saved a low six-figure sum. The company is also working to connect predictive maintenance alerts with its SAP Plant Maintenance system so that information identified by the monitoring software can be fed directly into the maintenance process.
I've seen the same pattern in restoration work. Equipment that has been in operation for decades carries a history of repairs, operating conditions, and design decisions. When a problem consistently arises, the question is why it's consistent and whether that finding should change a procedure, inspection standard, maintenance schedule, or design.
Manufacturing companies face this problem on a larger scale. Production has operational data. Maintenance has work orders and repair histories. Quality has inspection results. Engineering has design documentation. Operators have practical knowledge that may never enter a formal system.
NIST research found that different manufacturing decision-makers need different information, much of which comes from separate sources. The challenge is connecting these sources well enough for the company to learn from them.
Making Feedback Part of Continuous Improvement 📈
A feedback loop should connect an observation to an action and then measure what happened.
- If a maintenance interval changes, failures after the change should be tracked.
- If a production process is adjusted, quality and yield should be compared before and after.
- If a component is redesigned, its performance in operation should be fed back into the engineering record.
Bosch has used this approach in several of its plants. At its Hildesheim plant, the company used AI-based analytics during a production ramp-up and reported a 15% reduction in cycle times.
Bosch also used generative AI to create synthetic examples of production defects so that automated optical inspection models could be trained before enough real defects naturally occurred. For a stator inspection project, Bosch generated approximately 15,000 synthetic images. The company expects this method to shorten the project by six months and deliver annual productivity gains in the low six-figure euro range.
Documentation is also important. Throughout my career, I've learned that when something is documented, it can be repeated, evaluated, and improved. When knowledge resides with an experienced employee, the company relies on that person remembering it and being available. When the lesson is part of a documented process, others can test it and build upon it.
Deloitte's 2025 Smart Manufacturing and Operations Survey of 600 manufacturing executives revealed that while participants rated their human capital and maintenance maturity relatively low, they reported higher maturity in technology, operations, quality management, and continuous improvement.
Every production run, inspection, maintenance event, and process deviation gives a manufacturer another piece of information about how its operation behaves. The value comes when these lessons feed back into maintenance, engineering, production, and automation systems and change the next step.
The process is simple: observe what happened, understand why, make a change, and measure the result. Automation can accelerate this cycle and make it easier to repeat across a large operation. Learning still comes from what the company does with the information. When this loop is closed, automation becomes part of how the operation learns and improves over time.
