Researchers have developed a comprehensive dataset from consumer product recalls to enable data-driven risk assessment and enhance engineering safety decisions.

Key developments:

• Created a multimodal dataset from the CPSC recalls database to inform risk assessment using historical information
• Augmented the dataset with generative methods to improve utility
• Enables the application of computational methods and LLMs for analyzing product safety patterns
• Provides a foundation for proactive risk identification in engineering design processes

This research represents a significant advancement for engineering by transforming historical product failure data into actionable insights that can prevent future safety hazards, reduce recalls, and improve design decisions across industries.

RECALL-MM: A Multimodal Dataset of Consumer Product Recalls for Risk Analysis using Computational Methods and Large Language Models