Digitizing Fresh Food Supply Chains to Reduce Loss and Waste

Food loss and waste is an ever increasing challenge facing humankind from both a corporate and social perspective. A key driver of this challenge is the requirement to continue to offer and in some cases increase the range of products offered to the consumer both in and out of the local season. Coupled with this is also a requirement to increase, or at the very least maintain supply networks that are flexible, adaptive, responsive and fully transparent, comply with local and international policy and maintain tight margins. A solution requires the implementation of policies, procedures and frameworks to promote standardization, collective effort and shared value and burden across global food supply chain. To address these challenges, a digital remaining shelf life predictor tool was designed using an array state of the art Convolution Neural Networking (CNNs) and machine learning techniques architected within an array of low cost industry affordable technology. This digital tool is designed to (1) determine quality of and (2) predict remaining shelf-life (RSL) of fresh produce across the supply chain. Provisioning this information will empower all stakeholders regardless of their positioning in the supply chain, to adopt a First Expired First Out (FEFO) biologic-based supply chain, facilitating a product-based decision framework yielding more agile, responsive and efficient supply chains. Preliminary results indicate an overall accuracy of prediction of 99.9 % from a theoretical industry standard (best case scenario) of 84 %.