From Data Mining To Gold Mining

Rosina Geiger

The Acerta platform is unlocking the hidden value in production data and transforming it into actionable insight. It’s cutting manufacturer’s warranty costs by 30%—and saving millions of dollars.

The promise of Industry 4.0 and the smart factory is real. New digital technologies are helping manufacturers achieve greater flexibility, productivity, quality, speed, and safety at almost every production stage. This ability to adjust to data and learn from it in real time is driving a stark contrast between yesterday’s assembly line and tomorrow’s smart factory.

In this emerging landscape, manufacturers can now predict, prevent, and correct manufacturing defects that were invisible in the not-so-distant past. And that is dramatically improving performance while avoiding costly downtime.

Putting the old assembly line to the test

Traditionally, manufacturers have relied on one-dimensional pass-fail tests to ensure quality.

Testing stations are placed at critical locations on the manufacturing line. After measuring several technical parameters, the reading is scored and compared to a threshold tolerance. Based on that result, the product receives a grade of either Pass or Fail. If the product passes, it moves on to the next manufacturing stage. If it fails, it is sent to a technician for further analysis.

Unfortunately, a simple pass-fail analysis gives manufacturers a very incomplete picture of true quality. A car brake component, for example, may pass only to fail under extreme cold or heat. Similarly, a transmission model might pass with flying colors—unless it is observed in a complete ecosystem with other engine components. It is this holistic view of product quality and relationships that has eluded manufacturers—until now.

Unlocking the smart factory

Canadian startup firm Acerta analyzes manufacturing environments using machine learning and automotive expertise. By using intelligent algorithms that combine both advanced machine learning with technical logic, Acerta‘s software identifies useful correlations that go well beyond a simple pass-fail analysis. The result is a 360-degree product monitoring system that ensures the very best quality.

“As it turns out, the information about the problems–and the solutions–do exist deep inside the data,” says Acerta CEO and co-founder Greta Cutulenco. “Machine learning is a great tool for analyzing hundreds of different signals simultaneously and identifying non-linear, non-trivial, and complex interrelationships between them. When you provide engineers with such a tool, especially one that‘s designed for their specific tasks, you significantly increase their ability to identify hidden defects and perform quick and efficient root cause analysis. This leads, ultimately, to a 10 to 30 percent reduction in warranty rates.”

A platform for the future

Acerta’s SaaS platform’s ability to detect and predict failures in real time is a boon to engineers, allowing them to identify the root cause of problems up front. That saves millions of dollars in downtime, engineering resources, warranty claims, and legal expenses further downstream. For customers, it means a significant reduction in quality-related and safety-related expenses.

  • For a wind turbine manufacturer, the Acerta platform revealed that 40% of gearboxes that passed a standard EOL test were actually failing during the warranty period. By getting to the root cause, the customer was able to reduce gear tooth fracture warranty costs by €2 million per year.
  • Acerta helped a major automotive manufacturer ferret out a hard-to-find issue with engine misfires. This allowed defective engines to slip through conventional testing. But by using deep data analysis, Acerta discovered unexpected hidden relationships between exhaust gases and various engine signals. Had this problem not been detected, the engine would have stalled and potentially broken down mid-drive, which could have resulted in serious physical injuries or worse.

Going for gold

Since the founders first began their research, interest in Acerta’s platform among automotive industry leaders has exploded. Many manufacturers have already started using Acerta‘s systems to analyze data, detect, predict, and prevent system failures. These early quick wins have already made Acerta a reputable and trusted leader in machine analytics and predictive technology. “Acerta is transforming hidden problems into visible opportunities,” says Cutulenco. And that’s freeing automotive engineers to do what they do best–design and build great cars.”

The big idea

The Acerta SaaS platform can see hidden manufacturing defects using sophisticated data techniques. This is unlocking the hidden value in production data and transforming it into actionable insight.

Why it’s big

By using intelligent algorithms to discover hidden correlations among a car’s hundreds of sensor signals, Acerta’s holistic approach is reducing warranty cases by as much as 30 percent. This is saving manufacturers millions of dollars per year.

To learn more about how SAP is working with startups, get in touch with the SAP IoT Startup Accelerator and SAP.iO.


Rosina Geiger

About Rosina Geiger

Rosina Geiger is the Director of Startup Engagement at SAP. She has worked at the Hasso Plattner Institute in Potsdam before joining SAP in 2016 to establish the SAP IoT Startup Accelerator in Berlin and Palo Alto.