Automating food inspection using deep learning for Sunsweet Growers

Automating food inspection using deep learning for Sunsweet Growers

OneSix developed an AI-powered computer vision system for Sunsweet Growers, automating prune inspection with real-time defect detection while preserving high product quality standards.
Data Science
AI & Machine Learning

Overview

Automating the prune inspection process while maintaining quality

Sunsweet Growers, the world’s largest prune distributor, aimed to automate its century-old manual inspection process to ensure product quality at scale. The new solution needed to match the reliability of the traditional process and operate as an edge device in areas with limited connectivity.

Our client’s goal was to build a single integrated provider marketing strategy powered by artificial intelligence. To do so, we had to predict not only how likely individual providers are likely to engage, but also how to intervene to change their probability of engagement. Additionally, as one of the largest pharmaceutical companies in the world, the data volume meant our solution had to be robust and architected to operate at scale.

Our Solution

Building an computer vision system for real-time defect detection

OneSix developed a computer vision solution using deep learning, deployed on portable edge devices to inspect prunes on-site. The solution comprised three key components: a hardware device capturing images from multiple angles, a cloud-based computer vision pipeline analyzing images for defects using custom-trained models, and a data warehouse for storing and visualizing inspection data. The system included a ‘human-in-the-loop’ component, allowing annotators to provide feedback on defect analysis to continuously improve model accuracy.

Results

Reliable, automated inspection ensuring quality and scalability

Since deployment, the system processes millions of prunes each season, delivering high-quality products with real-time defect detection. Sunsweet Growers’ team relies on the integrated monitoring and reporting dashboard throughout each production season, benefiting from an efficient, scalable inspection process that preserves product quality.

“OneSix designed, built and deployed a solution that integrated computer vision models with a monitoring/reporting dashboard that our team relies on throughout each production season. Since its initial deployment, as new challenges and opportunities have arisen, OneSix remains a valued collaborative partner to Sunsweet.”

Ready to unlock the full potential of data and AI?

Book a free consultation to learn how OneSix can help drive meaningful business outcomes.