Zespri

Predicting fruit quality in the supply chain from harvest to market

Zespri

Problem specification

  • Two varieties
  • Orchards harvested over a range of maturities and different orchard have different quality and storage potential
  • Fruit shipped over a 6 month period
  • Orchards have to be shipped to arrive in the market in good condition – getting the shipping order optimized is the secret to a successful season
  • Knowing what lines to prioritise for shipping is only generally known. There are often lines that do not perform according to expectations and being able to predict which lines are unlikely to meet their potential (or more correctly, perhaps, accurately predict their potential) has potentially major benefits to the industry.
  • Features of this data are partial information, noisy data due to the fact information is based on small sample sizes

Possible outputs

  • New analytics and tools for analyzing complex fruit quality data
  • Improved designs for the sampling and monitoring systems
  • Optimisation of the quality prioitising process
  •  

Data available

  • Harvest quality parameters for all orchards in New Zealand
  • Monitoring data during storage
  • Quality information at various points along the supply chain