Sustainascan: Smarter Shopping For A Sustainable Future

Abstract

The rise in environmental concerns and consumer awareness has amplified the demand for sustainable shopping practices. However, identifying eco-friendly products based on their ingredients and environmental impact remains a challenge for everyday consumers. Traditional methods of assessing product sustainability require extensive research and lack real-time accessibility. The proposed work introduces SustainaScan, a web-based assistant that leverages Artificial Intelligence (AI) and Machine Learning (ML) techniques to automatically evaluate the sustainability of consumer goods. By utilizing image-based ingredient recognition and ECOSCORE computation, the system assesses carbon footprint, biodegradability, and toxicity levels to provide users with a clear sustainability rating. The platform also suggests environmentally preferable alternatives and redirects users to verified purchase sources. Various classification and scoring algorithms have been applied and tested for accuracy, precision, recall, and user impact. Despite the advancements, challenges persist in dataset completeness, OCR accuracy, and integrating real-time scanning capabilities. Future work aims to enhance model scalability and improve cross-platform deployment to broaden accessibility and effectiveness in promoting sustainable consumer behavior.
Index Terms—Sustainable Shopping, Artificial Intelligence (AI), Machine Learning (ML), ECOSCORE, Environmental Impact, Ingredient Recognition.