At TT PSC, we believe that manufacturing leaders should rely on reliable, real-time data to achieve the real goals of Industry 4.0 (r)Evolution. This data can build knowledge and awareness and identify actions requiring improvement.
In the Lesson Learned Explained series, we present examples of completed projects along with insights. We hope that this short material will help you avoid at least some of the pitfalls on the way to Digital Transformation.
Challenge: Enhancing Production with Data
An automotive company needed a solution to monitor production to improve the quality of final products. The key element was identifying issues by analyzing quality data correlated with production data. Special attention was given to the casting and cooling zones, where product quality was particularly variable.
Solution: System Integration and Data Analysis Using AI and ML
To achieve the desired goal, the company, with the help of TT PSC, implemented a solution that included integration with data from machines and sensors on the production floor. The key element was using an AI/ML engine with access to the process/plant measurements history. The collected data was visualized and analyzed for correlations to help identify issues affecting product quality. It allowed for real-time data monitoring and analysis, enabling quick responses to any irregularities.
Implementation Challenges: Lack of Definitive Link Between Process Data and Quality
Despite advanced data analysis, the solution did not find a definitive link between process data and product quality. It was discovered that the problem was not the production process itself. Statistical tools (ML) identified a correlation with a parameter that seemed unrelated to the process – the HVAC system outlet. It turned out that the direct airflow from the cooling system affected product quality. This discovery was crucial as it identified the real source of the problem, which was not directly related to the production process.
Adjustment: Changing Cooling Conditions in the Production Area
The solution to the problem was simple: shield the cooling area from direct airflow. The goal was to avoid the impact of air conditioning on the cooling process of castings, which was intended to stabilize the quality of final products. This change made it possible to eliminate quality variability caused by uncontrolled cooling conditions. This simple solution brought significantly improved product quality.
Result: Effective Identification and Resolution of Quality Challenges
The client’s experts had long struggled with unstable casting quality and needed help finding the cause. After solving it, it became clear that there was no link between process data and production quality. Only the application of statistical analysis (ML) to all available parameters allowed the identification of a correlation with a seemingly unrelated parameter—air conditioning. Quality degradation occurred when the air conditioning was on, affecting the heating/cooling curve.
Conclusions
Implementing advanced AI/ML technologies in monitoring the production process allowed for the identification and resolution of quality issues that were difficult to detect using traditional methods. System integration and real-time data analysis enabled quick responses to irregularities, significantly improving the quality of final products. The key to success was the collaboration of industry experts with advanced analytical tools, highlighting the importance of synergy between human knowledge and technology.
Based on the actions taken, several key conclusions can be drawn. First and foremost, algorithms should support the decision-making process as they can analyze data that would typically be overlooked by an expert’s “intuition.” The knowledge of industry experts and production specialists is essential for effectively using machine learning. Even complex algorithms should provide results that are easy to understand and practical. This makes it possible to monitor and improve product quality in the automotive industry effectively.
_Are you interested in digital transformation for your enterprise?
Check out other implementation examples below and the lessons learned from them.
