Research by BSES into the applications of component analysis by near infra-red
retlectance (NIR) over more than 10 years has culminated in the use of the technique for clonal evaluation of sugarcane in plant breeding trials. Researchers have also examined the analysis of mill prepared cane in an at-line situation. The extension to on-line analysis of prepared cane, a high moisture content material, was a development new to on-line technology, particularly in the sugar industry. This paper reports work which has led to on-line scanning of prepared cane and use of the technique in individual rake fibre estimation in cane payment at Mulgrave Mill. The use of a Foss NIRSystems
6500 system is described. It consists of a scanning head, mounted adjacent to the feed
chute on number one mill, scanning a continuous stream of processed material and
linked to a near infra-red spectrophotometer which included a monochromator for
resolving the reflected light into light of a discrete wavelength. The spectrophotometer
was in turn linked to a computer loaded with the database to convert spectra to an
analysed result, to manage the analytical process and to interface with the mill's cane
payment and process control computers. The development of calibration equations by
gathering spectra for comparison with routine laboratory analyses, the process of
implementation and outcome in use of calibration equations for fibre is described. The
potential of the system and initial outcomes for parameters other than fibre including
juice brix, juice polarisation, commercial sugar content (CCS), water, ash, pol in open
cells, inorganic substances and extraneous matter are also presented.