Luke Qian
Ph.D. Candidate, Wiedmann Lab, Cornell
University
Transforming AI Models into Applications for Improved Food Safety and
Quality
Abstract
With advances in artificial intelligence (AI) technologies,
the development and implementation of digital food systems are becoming
increasingly possible. There is tremendous interest in using different AI
techniques, such as machine learning, natural language processing, and computer
vision to improve food safety and quality. Possible AI applications are broad
and include, but are not limited to, 1) food safety risk prediction and
monitoring as well as food safety optimization throughout the supply chain, 2)
improved public health systems (e.g., by providing early warning of outbreaks
and source attribution), and 3) detection, identification, and characterization
of foodborne pathogens. The utility of AI models for improved food safety and
quality is illustrated with three projects that aim to develop 1) a machine
learning model that predicts the spore level in the raw milk on the farm level,
2) a Monte-Carlo simulation model that predicts the percentage of milk
containers spoiled due to psychrotolerant sporeformers, and 3) an agent-based
model that predicts the Listeria contamination on the surfaces of a retail
environment. However, the potential of these digital tools depends on the
availability and quality of data, and several obstacles need to be overcome to
achieve the goal of digitally enabled “smarter food safety” approaches.
Existing solutions for improving data sharing, while not compromising data
privacy, are discussed, most notably differential privacy and federated
learning. In addition, user-centric deliverables that match the business
interest of relevant stakeholders are needed to promote the commercial use of
developed digital tools.