Next-Generation Smart Campus Dining: A Survey of Real-Time Resource Allocation, LLM-Driven Menu Optimization, and Access Control Paradigms

The integration of real-time web technologies and Generative Artificial Intelligence (AI) into campus infrastructure is creating new paradigms for institutional resource management. In campus dining environments, traditional offline management leads to significant food waste, inefficient resource allocation, and delayed feedback loops. This survey reviews the transition from static Enterprise Resource Planning (ERP) systems to dynamic, AI-optimized ecosystems. We synthesize recent advancements in multiple core areas: (i) frictionless Role-Based Access Control (RBAC) using decentralized access codes, (ii) real-time data synchronization for live inventory telemetry via WebSocket protocols, (iii) the application of Large Language Models (LLMs) for automated, sentiment-driven menu generation using structured prompt engineering, (iv) mathematical modeling of sustainability metrics to reduce organic waste, and (v) fault-tolerant microservice architectures. By bridging the gap between operations management and student end-users, this paper outlines how modern architectural frameworks—such as those utilizing Node.js backends, document- based NoSQL databases, Redis-backed event brokers, and Gemini AI integration—can resolve the critical information lag in institutional dining. We propose the TasteTrack architecture as a comprehensive, scalable solution to these challenges, detailing its mathematical foundations, cryptographic privacy models, and algorithmic workflows.