BASF Automotive Refinish Solutions has developed a software program that applies artificial intelligence using neural networks to the process of color matching paint.
At this time, BASF’s artificial intelligence color matching system is only used internally. In the future, BASF plans to offer this system to its global automotive refinish customers and leverage the technology in the OEM coatings market.
Developed by BASF’s Craig McClanahan, manager of color research for BASF Coatings North America, the artificial intelligence improves the efficiency and accuracy of the color matching process, BASF says. The software uses a machine learning algorithm to achieve its color matching ability from previously developed color formulations.
“One of the significant elements of this artificial intelligence is that it takes color matching to a higher level with a greater degree of accuracy than ever before seen in the automotive refinish industry,” said McClanahan. “It actually learns and gets better over time.”
As BASF produces new color matches over time, the intelligent system learns from them, then automatically and exponentially applies that learning to the matching of new colors. BASF claims the new system achieves high-quality color matches 95 percent of the time, versus the industry standard of roughly 50 percent.
McClanahan and the artificial intelligence development team located in Whitehouse, Ohio, use information from BASF’s globally integrated color laboratories around the world to “train” the artificial intelligence software with global BASF color-matching information. By the end of 2009, this artificial intelligence will contain color data from all 10 BASF automotive refinish laboratories in Japan, Australia, China, Brazil, France, Germany, India, Spain, Italy and the United States.
For more information, visit www.basfrefinish.com.
