Design of an AI-Powered Smart Shopping Cart with Inclusive Accessibility Features

This paper presents the design and development of an AI-Powered Smart Shopping Cart with inclusive accessibility features tailored for supermarkets in Peru. Supermarkets face challenges to improve accessibility, reduce wait times, and provide personalized customer experiences, particularly for people with disabilities. In Peru, these issues are exacerbated by infrastructural limitations and a lack of technological solutions in retail. To address this, we introduce a system that includes an AI virtual assistant utilizing prompting for natural and personalized user interactions, AI-driven product recognition via cameras, and a height-adjustable mechanism designed specifically for wheelchair users. The product recognition system uses YOLOv8obb for barcode detection and a custom convolutional neural network architecture for barcode reading. This system not only improves accessibility, but also achieves a precision rate of 97.8% in real world environments, surpassing existing solutions. This design represents a significant advancement in the integration of technology and accessibility in retail environments, offering substantial benefits for both consumers and supermarkets.