The Research Institute for Computational Sciences at the National Institute of Advanced Industrial Science and Technology has developed a new technology for the rapid and accurate measurement of polymer density, the first technique of its kind in the world. The research was conducted jointly with Toyama Industrial Technology Center and Opt Research Inc. The assay method is expected to have applications in material recycling of waste plastics and to facilitate production of high quality polymers.
The research group combined near infrared reflection spectra measurement and neural network analysis to produce a method that could provide more rapid results than existing plastic density assays. Near infrared spectra were measured for polyethylene samples of known densities ranging between 0.90 and 0.96 g/cm3, and the spectral data were entered into a neural network.* This method is the first in the world to produce rapid results on polyethylene density to within a 0.2% margin of error.
* Neural network: Information processing technology that emulates the signal processing performed in nerves in the human brain. The technology is used in economics and other disciplines, as well as in scientific fields. The development of brain-like computers is now being targeted.