In the rapidly evolving landscape of AI and e-commerce, Kutluhan Aktar, an Engineer from Turkey, has created an AI-driven IoT shopping assistant with ChatGPT. This innovative system aims to merge the digital convenience of online shopping with the tactile experience of in-store shopping. Unlike the high-cost self-checkout solutions such as Amazon Go, Aktar's invention is tailored for local businesses, offering them a cost-effective way to integrate AI-driven e-commerce features without the need for extensive renovations or high monthly cloud service fees.
The AIoT shopping assistant is built upon a robust hardware foundation. Central to its operation is the WIZnet W5300 TOE SHIELD, which ensures high-speed bus communication and reliable Ethernet transmission speeds. This shield is paired with the STM32 NUCLEO-F439ZI development board, programmed in MicroPython, to form the heart of the device. To facilitate the scanning of barcodes and QR codes, a GM77 barcode and QR code scanner is integrated. For user interaction and feedback, an SSD1309 transparent display and an RGB LED are incorporated. Additionally, a DFRobot Tiny (Embedded) Thermal Printer provides instant printouts, enhancing the user experience. The entire setup is powered by a Xiaomi 20000 mAh 3 Pro Type-C Power Bank, ensuring longevity and reliability. The hardware components are housed in a specially designed checkout counter-themed case, which not only protects the components but also offers a user-friendly interface for shoppers.
The core of this system is a robust e-commerce web application, meticulously crafted using PHP, HTML, JavaScript, CSS, and MySQL. This application is designed to communicate with a physical shopping assistant device. By scanning barcodes, the most universally recognized product identification method, it fetches product details and offers AI-generated shopping recommendations to the user. This seamless integration ensures that customers can effortlessly create accounts, scan products to add or remove them from their cart, view their cart's current status, receive AI-powered product suggestions based on their selections, and finalize their purchases.
One of the standout features of Aktar's project is its design. The shopping assistant device features a checkout counter-themed case, offering a familiar touchpoint for customers. The interface is user-centric, ensuring ease of use and a pleasant shopping experience. Moreover, the system uses the OpenAI API to generate ChatGPT-powered recommendations and the Open Food Facts JSON API to obtain product information. It also integrates Brevo's Email API to send HTML emails directly from localhost, keeping the user informed about their shopping activities.
