Artificial Intelligence for Paint Shops
- PDF / 282,474 Bytes
- 2 Pages / 595.276 x 790.866 pts Page_size
- 83 Downloads / 448 Views
Artificial Intelligence for Paint Shops A new artificial intelligence application for paint shops identifies the sources of defects and specifies ideal maintenance schedules. The first practical applications show that the software increases plant availability and improves the surface finish of the painted vehicle bodies.
Why does the same defect occur unusually often on one part of a vehicle body? When is the latest time at which a mixer in a robot can be replaced without causing a machine stoppage? Precise answers to these questions can make an important contribution to long-term financial success, because every defect and every unnecessary maintenance task that can be avoided will save money or increase product quality. In the past there has been a lack of accurate information that could be used to detect faults or failures at an early stage. Any data that was available was generally the result of painstaking manual evaluations or trial and error. Artificial intel-
ligence (AI) allows much more precise information to be generated automatically.
The special feature of the new software package is that it combines large volumes of data, including historic data, with machine learning. It could be said that the selflearning AI application has its own memory. This enables it to identify complex correlations in large amounts of data on the basis of past information and to predict a future event to a high degree of accuracy by evaluating the current status of a machine. The software can be used for a wide range of applications in paint shops on the component, process and plant level.
AI application with its own memory The new Advanced Analytics self-learning plant and process monitoring system from Dürr is the latest member of the DXQanalyze product family, which already includes Module Data Acquisition for acquiring production data, Visual Analytics for visualizing it and Streaming Analytics. Using a low-code platform, Streaming Analytics allows plant operators to analyse almost in real time whether the production process is deviating from previously defined rules or target values.
Predictive maintenance reduces downtimes
© Dürr
When it comes to components, Advanced Analytics aims to reduce downtimes by means of information about predictive maintenance and repairs, for example by predicting the remaining service life of a mixer. If the component is replaced too early, there will be an unnecessary increase in spare parts costs and maintenance work. However, leaving it too long to replace the part can result in a deterioration in the quality of the coating and machine stoppages.
12
IST International Surface Technology 4 I 2020
The new AI application identifies the sources of defects and specifies ideal maintenance schedules.
© Dürr
Using artificial intelligence, systematic defects in the painting process can be tracked down and the overall effectiveness of the plant can be increased.
Using high-frequency robot data, the software learns to identify wear indicators and wear patterns over time. Because the data
Data Loading...