Chapter 5 - A Data-Driven Approach to Food Safety Surveillance and Response

Abstract

Increasingly, new techniques of big data and predictive analytics are being marshaled to reduce the time, scale and scope of foodborne contamination events. Contamination of food can occur at any point across increasingly complicated and intersecting global food chains. By the time a food safety problem is suspected, several weeks may have passed since first contact with the tainted product. And only after a foodborne outbreak has been confirmed by laboratory tests can the responsible products be identified and recalled. The entire process from detection to product recall can take weeks, even months. Dealing with the twin problems of early detection and rapid response are the core challenges of food safety today. In this chapter we describe a prototype informatics tool called NCFEDA (North Carolina Foodborne Events Data Integration and Analysis) that builds situational awareness of emerging contamination events by fusing traditional and nontraditional data sources, predictive analytics, visualization tools, and real-time collaboration across stakeholders to reduce the latency in detecting and responding to emerging contamination events.