Improving Nutritional Intake Tracking with an Intelligent System: Designing an Integrated Nutrition Framework and Prototype

Abstract

Current nutrition technologies often use separate tools for food logging, recipe discovery, activity tracking, and dietary analytics. This separation makes it hard for users to see nutrition as a connected process, leading to confusion, inconsistent data, and low engagement over time. Inspired by enterprise resource planning (ERP) systems, this thesis suggests that nutrition should be managed as an integrated, lifecycle-based system instead of as isolated tasks. This study uses a Design Science Research approach and makes two main contributions. First, it introduces the Integrated Human Nutrition framework, which models nutrition management as three connected cycles: food production, food consumption, and activity. Each cycle follows clear Input, Process, and Output steps. Second, the study describes the design and creation of a working prototype. This web-based system brings together food acquisition, pantry management, recipe generation, meal logging, and progress tracking in one place. The prototype uses a React frontend, a FastAPI backend, and a PostgreSQL database, along with external nutrition and authentication services. The system is evaluated by assessing how well it matches the intended concepts, testing it in real-world scenarios, and comparing it with a commercial nutrition app. Results show that the prototype successfully puts the consumption cycle into practice, reduces the fragmentation of nutrition tasks, improves users' awareness of their eating habits, and is consistent and easy to expand. While the prototype is not positioned as a commercial alternative, the evaluation indicates that an ERP-inspired, lifecycle-aware design provides clear structural advantages over siloed nutrition applications. This research lays the groundwork for future system improvements, real-world testing, and wider use of integrated information systems in nutrition informatics.