The Innovation Navigator: A Prototype for AI-Augmented Design Thinking

Abstract

This thesis investigates the practical integration of artificial intelligence (AI), particularly large language models (LLMs) and related technologies, into structured innovation processes, specifically design thinking. Despite growing theoretical interest and technological advances, AI adoption in innovation management remains fragmented, with existing tools offering limited, isolated support disconnected from established methodologies. Addressing this gap, this research conceptualizes, develops, and tests the "Innovation Navigator," a prototype for an AI-augmented software tool that operationalizes AI capabilities, such as generative language models, retrievalaugmented generation, web crawling, and image synthesis, within a modular, humancentered innovation framework grounded in design thinking and related innovation methods. Building on previous exploratory work and empirical input from startup practitioners, the Innovation Navigator embeds AI functionalities into well-known innovation templates, including personas, empathy maps, and business model canvases, facilitating a structured yet flexible innovation journey across all innovation phases from organizational alignment to knowledge preservation. Utilizing the Design Science Research Methodology (DSRM), a proof-of-concept prototype was developed and qualitatively evaluated through user testing and semi-structured feedback sessions with entrepreneurs in early-stage startups. Besides accelerating the innovation process and complementing human capabilities through generating insights and suggestions, the Innovation Navigator addresses common hierarchical, psychological, and cognitive barriers in design-thinking workshop settings by enabling anonymous ideation, promoting experimentation, and expanding creative horizons. This thesis offers a practical tool and critical first proof-of-concept of human-AI collaboration in innovation management, demonstrating the feasibility and value of AIaugmented design thinking tools. It outlines directions for future empirical validation, tooling refinements, and broader application scopes, ultimately aiming to advance cognitive automation and hybrid intelligence in innovation practice.

Date of Award	2025
Original language	English
Supervisor	Kristiana Roth (Supervisor)