AI-Powered Self-Service Outdoor Kiosk (MVP)

1. Introduction SOLUTION NEEDED IS MVP 1.1 General Project Description The project aims to develop advanced software for an AI-powered self-service kiosk designed for outdoor environments. It is optimized for challenging conditions such as ambient noise and direct sunlight. This kiosk enables seamless and personalized customer interaction via voice commands or a touchscreen, ensuring an efficient and optimized shopping experience. Technologies to be used change with the developer skillset. Operational Conditions: Every interaction (proposals, promotions, etc.) must include a Yes/No confirmation message. Responses must be immediate. Offline performance is required. Text descriptions of the BOT's speech should be displayed on-screen. Clear visualization of UI components. The shopping cart total must always be visible on-screen. A navigation bar must always be available. Key Features: Fully sealed kiosk with AI technologies to: Process requests and manage an interactive shopping cart. Offer personalized recommendations based on user and environmental conditions. Support multiple payment methods for completing purchases. 1.2 System Objectives Deliver a fast, personalized, and innovative shopping experience with seamless voice or manual interaction. Reduce operational costs and maximize efficiency in small spaces. Ensure continuous offline operation with automatic synchronization when online. 1.3 Special Features Outdoor Operation: Built for external environments, resistant to external noise, weather, and vandalism. Hybrid Interaction: Switch between voice and touch interaction. Voice-Controlled Shopping Cart: Customers can manage their cart (add, remove, or check products) using voice commands. 2. Hardware Requirements Main Unit: Processor: Intel i5 or higher (mid-to-high range). RAM: Minimum 16 GB. Storage: 256 GB SSD. Programming Technologies (To be finalized with developers): Front-End: React, HTML, [login to view URL], Angular. Back-End: C#, FastAPI. Database: MySQL, SQLite. Input Devices: Palm recognition for payments. Biometric fingerprint scanner. 3D camera for facial recognition. Omnidirectional microphone with noise cancellation. Sunlight and water-resistant touchscreen (IP65 or higher). Proximity sensor. Output Devices: Speakers for synthesized voice output (TTS). Ticket and receipt printer. Additional Peripherals: Barcode and QR code scanner. NFC reader for card payments. Palm and facial payment readers. Wireless credit/debit card reader. Temperature sensor for contextual recommendations. 3. System Features 3.1 Modules and Services BOT 1: Welcome & Identification BOT Purpose: Identifies users through biometrics (face, fingerprint, palm), credentials, or other means. Provides a personalized, empathetic welcome based on user data, such as hobbies and interests, and environmental data like weather or news. Key Features: Identify customers via facial recognition. Provide a personalized and empathetic welcome using TTS and on-screen text. Contextual updates (e.g., sports scores for fans). Integration with APIs for weather, local news, and customer databases. BOT 2: Promotion BOT Purpose: Gathers cart data, consults the store database for prioritized suppliers or brands, and creates tailored promotions. Key Features: Real-time promotions based on cart content. Pair product offers (e.g., "food + drink" combos). Sponsored product recommendations. Interactive promotion pop-ups. BOT 3: Upselling BOT Purpose: Suggests additional items based on purchase history, store trends, and impulse-buy products like chocolates or gum. Key Features: Personalized upselling using purchase history. Recommendations for popular items and impulse products. TTS-based promotions. 3.2 Voice Interaction Voice Capture & Processing: Robust speech recognition (e.g., Whisper). Intent identification via AI (e.g., GPT or similar). Voice Responses: Natural voice output (TTS). Fluid conversations that recall customer data. Cart Management via Voice: Commands for adding, querying, and removing items from the cart. Example: "Add a cola to the cart" or "What’s in my cart?" Environmental context for suggestions (e.g., cold drinks on hot days). 3.3 Product and Cart Management Database Integration: Local database for product availability. Real-time inventory updates post-purchase. Advanced Cart Features: Add, modify, or remove items via voice or touch. Auto-calculation of totals, taxes, and discounts. Always display the cart and total on-screen. 3.4 Manual Options Intuitive touchscreen GUI for product browsing and cart management. Predictive suggestions in the user interface. Smooth transition between voice and manual interaction. 3.5 Sales and Payments Sales Process: Cart confirmation by the customer. Digital or printed receipts. Payment System: Integration with Transbank for card payments. Support for QR payments, digital wallets, biometrics (palm, fingerprint, face). 4. Database Model (SQLite/MySQL/PostgreSQL) Tables include: customers: Stores customer data (e.g., ID, name, hobbies, biometric data). products: Details of store products (e.g., ID, description, stock, expiration, price). sales: Tracks individual product sales linked to a transaction ID. historical_sales: Summarizes total sales per transaction. store: Information about the physical store (e.g., location, IP). promotion: Fixed discounts for products. campaign: Tracks active campaigns and prioritization for upselling. 5. Scalability and Operation Offline & Online Modes: Full offline functionality with automatic synchronization when online. Local database ensures uninterrupted operation. Future Integrations: Additional payment methods. New product categories (pharmacy, pet supplies). Advanced recommendations with expanded data. 6. Operational Benefits Cost Reduction: Minimizes staffing needs. Maximizes efficiency in compact spaces. Technological Innovation: Enhances customer experience with advanced personalization. Revenue opportunities through recommendations and ads. Sales Optimization: Increases average ticket size via upselling. Rapid product search and cart management for quick transactions.