Designing a User Interface for Generative Design in Augmented Reality: A Step Towards More Visualization and Feed-Forwarding

2025-03-27

AI Summary

Current AR environments lack efficient, user-friendly generative design interfaces, hindering their adoption in immersive design scenarios. To address this, we present the first AR-native generative design UI system, which innovatively enables automatic constraint inference from user input, dynamic virtual environment construction, real-time object editing, and structured CSV export. Grounded in participatory design principles, the system integrates anticipatory guidance with immediate visual feedback. Technically, it unifies AR rendering, generative algorithms, constraint reasoning, and visualization engines to support end-to-end immersive interaction. Empirical evaluation demonstrates significant improvements in design operation efficiency, intuitiveness, and task completion rate. Our work establishes a reusable interaction paradigm and foundational technical framework for native generative design in augmented reality.

Technology Category

• Humans and AI: Game Design — Procedural Content Generation & Storytelling
• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Cognitive Modeling & Cognitive Systems: Computational Creativity

Application Category

• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems

Abstract

Generative design, an AI-assisted technology for optimizing design through algorithmic processes, is propelling advancements across numerous fields. As the use of immersive environments such as Augmented Reality (AR) continues to rise, integrating generative design into such platforms presents a potent opportunity for innovation. However, a vital challenge that impedes this integration is the current absence of an efficient and user-friendly interface for designers to operate within these environments effectively. To bridge this gap, we introduce a novel UI system for generative design software in AR, which automates the process of generating the potential design constraints based on the users' inputs. This system allows users to construct a virtual environment, edit objects and constraints, and export the final data in CSV format. The interface enhances the user's design experience by enabling more intuitive interactions and providing immediate visual feedback. Deriving from participatory design principles, this research proposes a significant leap forward in the realms of generative design and immersive environments.

Problem

• Lack of efficient AR interface for generative design
• Need for intuitive user interaction in AR design
• Automating design constraint generation in AR environments

Innovation

• AR-based UI for generative design automation
• Interactive virtual environment with real-time feedback
• CSV export for design constraint data