Designing AI Utilization Processes for Brand Identity Design = 브랜드 아이덴티티 디자인을 위한 AI 활용 프로세스 설계

다국어 초록 (Multilingual Abstract)

Research Background・Purpose
Brands are a key element in conveying a company's identity and values ​​to consumers, and brand identity design is utilized as a means to enhance a company's competitiveness and establish a positive image in the market. Recently, AI(Artificial Intelligence) technology has been utilized in various fields requiring creativity, and research on its application in brand identity design is also underway. Recent research primarily focuses on AI logo design platforms using the "Wizard Process" method and the image generation capabilities of generative AI. While AI logo design platforms offer advantages in accessibility and efficiency, their reliance on user input through a fixed system can limit design quality and creativity. In contrast, generative AI possesses autonomous design generation capabilities, enabling the creation of diverse design styles. However, it requires a prior understanding of the user's brand concept and has limitations in producing clear design results. Compared to these characteristics, designer-focused traditional approaches offer easier access to sophisticated design results through brand strategy development. However, they limit idea expansion and require significant time. Accordingly, this research aimed to complement the functional limitations of AI discussed in previous studies, analyze AI's design generation mechanism from a brand identity perspective, and then combine it with a designer-focused traditional method to design an efficient design process utilizing AI tools. In the context of this research development, this research set the following research questions.

Research Questions
RQ 1. What is an efficient design process for utilizing AI tools (AI logo design platform・generative AI) for brand identity?
RQ 2. Are there differences in users’ usage and perception of AI tools (AI logo design platform・generative AI) depending on their level of expertise?
RQ 3. What are the principles of BI design produced by AI tools (AI logo design platform・generative AI)?
RQ 4. What are the effects of AI utilization in terms of creativity in BI design?

Research Methods
The research process and methods are as follows. In Step 1, the researcher established a designer's traditional brand identity design process by collecting and analyzing previous research, domestic brand identity design cases from the past five years, and process data from a design agency. In Step 2, the researcher analyzed the top eight AI logo design platforms in Google Search, designed an AI logo design platform process by combining it with the traditional process, and verified it through user research. In Step 3, the researcher conducted multiple brand logo design experiments using an AI learning method and combined them with the traditional process to design a generative AI learning process. User research was conducted to verify the efficiency of the process, explore the potential for application design, and propose a process that subdivided the roles and collaboration methods of designers and AI. In Step 4, the researcher applied the designed AI learning process to an actual corporate project, from strategy development to application system implementation, recording and analyzing the details. After the project was completed, a survey was conducted on the perception of the design results. This established the practical applicability of the AI ​​learning process and the relationship between designers & AI.

Conclusion
Within the context of the previous research, this study was conducted around four main research questions (RQ1-RQ4). To systematically respond to each research question, a theoretical framework was established, and an analysis of the functional characteristics and efficiencies of AI tools was conducted from the perspective of their utilization. Additionally, by investigating the differences in users’ usage and perception of AI tools depending on their level of expertise, the researcher examined the various approaches that emerge in the process of utilizing AI. In addition, the researcher analyzed AI's design generation mechanism and the visual patterns of the resulting design results, and closely observed how AI contributes to the creativity of BI design through practical projects that applied iterative experiments and processes. Through this, the research results corresponding to each research question were organized as follows.

First(RQ1), Regarding the efficient design process utilizing AI tools, the AI ​​logo design platform offers the advantage of quickly producing design results through wizard-based, easy setup. However, its limited automation system limits its ability to fully reflect the designer's strategic thinking and step-by-step design execution. Accordingly, this study planned an effective design process design that can reflect the designer's expertise. Specifically, by segmenting the industry selection category during the user information input stage, expertise was enhanced in the initial setup stage. Next, the user configured it so that users can set key visual elements such as concept keywords → design type → font, and color in stages, ensuring clarity in setting the design direction. Lastly, based on the input values set in the design creation stage, AI automatically generates multiple design drafts, and the user selects, compares, and finally selects a preferred design among them, thereby increasing the efficiency of design result selection through a systematic approach structure. Therefore, the efficient design process of the AI logo design platform is to strengthen the expertise of the initial setup for users and design a systematic system that increases the usability of the design development process, thereby enhancing the expertise and decision-making effectiveness of BI design. Generative AI(DALL-E) is based on autonomous image generation technology and has potential applications in a wide range of fields. However, unlike the AI logo design platform, it is not specialized for BI design, limiting its ability to produce design results that reflect the unique characteristics of a brand. Consequently, this study designed an organic design process that would incorporate the expertise and thought processes of BI design into the generative AI's operations step by step. To achieve this, data such as the brand's background, concept, and visual standards were input. A design exploration process based on this data was then conducted to guide the generative AI's operation in a direction consistent with the brand context. Next, the collaborative loop was strengthened through repeated prompt adjustments and result interpretation to gradually converge on the brand's visual exploration scope. Finally, the mutual exchange of opinions between the designer and the generative AI resulted in improved design results. Therefore, an efficient design process utilizing generative AI is to form a creative collaborative system where designers and AI organically interact based on 'BI design expertise' and gradually develop and converge BI design results.

Second(RQ2), User surveys of the AI logo design platform and the generative AI revealed differences in users’ usage and perception of AI design tools depending on their expertise. Non-designers tended to use the AI logo design platform by accepting the provided options and automatically generated results, and were more satisfied with ease of access and ease of use than with the design's completeness. This trend was similar in the generative AI, where greater significance was placed on the experience of using the generative AI than on the results of the creation itself. Conversely, designers focused on the speed of design creation and the way visual elements were presented based on option settings when using the AI logo design platform. With the generative AI, they placed greater value on the process of changing the design output through prompt adjustments. These differences suggest that users have different usage purposes and expectations for AI tools depending on their expertise level. Therefore, these user characteristics should be taken into account when designing AI tools for BI design.

Third(RQ3), Analysis of the design generation mechanisms and visual patterns of design results from the AI logo design platform and the generative AI revealed that the two AI tools generate BI designs based on different operating principles, which also presented limitations. The AI ​​logo design platform operates on the principle of repeatedly combining and presenting existing stored database designs according to an algorithm. User-entered information such as logo shape, color, and font are used to limit the design scope. The AI ​​then selects images that match these criteria from the database and combines them with brand name text in a limited system font to generate the design. Afterwards, users refine the final design using the limited editing features provided by the platform and purchase it. This template-based design approach can be effective in ensuring consistency in design results and work efficiency. However, it has limitations in ensuring originality and differentiation in BI designs, and problems can arise in the process of securing rights, such as trademark registration. Meanwhile, the generative AI operates on the principle of predicting and suggesting visual elements suitable for BI design based on user-entered prompts, based on a large amount of diverse data learned. The generative AI, based on a mechanism that converts linguistic information into visual representations, interprets the meaning contained in brand names, concepts, and directives, and then probabilistically combines related typography and symbolic images to generate design results. Compared to the AI ​​logo design platform mentioned above, this generation method has no fixed rules or structure in the generation process, and shows more flexible design performance in expanding the user's intention and combining the spelling structure of the brand name with visual images. However, the generative AI can significantly vary in the accuracy of its results depending on the linguistic interpretation of the input prompt, and it also exhibits limitations in repeatedly generating specific styles or visual elements. Furthermore, generative AI, which is based on image combinations based on existing data learning, may face trademark rights issues, such as the AI logo design platform, if it is not accompanied by a designer's post-processing process.

Fourth(RQ4), A close examination of AI's contribution to the creativity of BI design revealed that the AI ​​logo design platform and the generative AI demonstrated different utilization effects. The AI ​​logo design platform, through its repeated presentation of visual results, provided designers with a foundation for comparing, reviewing, and establishing the visual direction of BI designs. This characteristic can serve as a supplementary tool, reducing trial and error in the designers' initial idea generation phase by structuring the design selection and comparison process. Therefore, from the perspective of creativity in BI design, the effectiveness of utilizing the AI logo design platform lies in increasing the efficiency of setting initial design directions by providing designers with creative judgment criteria. However, designers' involvement is limited to the final stages of design results, limiting their ability to reflect the inherent creativity of humans through diverse experimentation and exploration. The generative AI has been shown to effectively expand designers' creative thinking and explore new ideas by engaging them step-by-step throughout the entire thought process through interaction with them. During the rebranding development process through collaboration with AI, the researcher deepened their understanding of the brand mission and concept through interaction with AI from the initial strategy stage. They then categorized, selected, and collected relevant keywords and ideas, input them into the AI, and attempted to combine different visual elements. Through these repetitive actions, the researcher captured clues for new design expressions and applied and developed those clues into actual brand designs. This demonstrates that combining a designer's unique creativity with AI technology can lead to effective design results. Therefore, from a creativity perspective in BI design, the effectiveness of generative AI lies in expanding designers' thinking and facilitating creative judgment through a co-learning process with them.

In a context where AI literacy and designers' collaborative abilities are increasingly required, this research aims to contribute to the field of brand identity design based on the following research findings. First, the the AI logo design platform research diagnosed the limitations of existing online platforms and proposed a design method that reflects the expertise of designers, thereby suggesting the possibility of the AI logo design platform providing expertise and strategic value. Next, the generative AI research proposed a method to expand designers' creative thinking and strengthen their problem-solving capabilities by designing a creative structure(co-learning) in which designers and AI organically collaborate. Finally, by applying this process to actual corporate projects and analyzing the collaboration methods, role division, and factors that facilitated thinking expansion between designers and AI. This analysis suggested a direction for human creativity and automated creativity to operate in a complementary manner. Through this, this research is significant in that it systematizes practical methods for future designers to utilize the efficiency and creativity of AI in a balanced way.

Suggestions
The design field is entering an era of co-creation, where designers and AI collaborate. This research analyzed the characteristics of the AI logo design platform and the generative AI and explored ways to increase their efficiency for brand identity design. Through this, the researcher aims to redefine the role and thinking of designers and to discuss changes in the future brand identity design structure. Previous research has demonstrated that AI logo design platform and generative AI each promote creativity in different ways. In this context, the roles and thoughts of designers who encounter each type of AI are summarized as follows. First, in utilizing the AI logo design platform, designers must strengthen their exploratory attitude to analyze the visual patterns and structures of AI results, seek creative clues, and recognize themselves as curators capable of selecting possibilities within the results. The generative AI emphasizes the designer's proactive ability to reflect and adjust various elements of clear intent, realism, and concept within prompts. Beyond controlling AI, designers must design the AI ​​generation process itself and broaden their depth of thinking to address problems. Finally, this researcher would like to emphasize that AI is facilitating changes in the way brand identity design is produced and created. Brands are evolving into complex systems that reflect online-centric consumer values and lifestyles. From this perspective, online-based AI can rapidly convey evolving consumer sentiments and culture to designers. Therefore, the future brand identity design production structure will be reorganized into an online creative platform that combines AI automation with designers' real-world contexts, acting as a creative medium connecting designers with the language of the times. However, advanced brand identity design is still achieved through human communication, and it is expected that AI will have difficulty completely replacing the areas of professional designers, such as strategic judgment, realistic brand identity design, and typography implementation.

목차 (Table of Contents)

• LIST OF FIGURES
• LIST OF TABLES
• ABSTRACT
• 1. Introduction
• 1-1 Research Background and Purpose
• 1-2 Research Methods
• 2. Theoretical Framework of Brand and Design Process
• 2-1 Brand Definition
• 2-2 Understanding Brand Identity Design
• 2-3 Research on the Processes
• Previous Research on Design Processes
• Previous Research on Brand Identity Design Processes
• Establishing a Designer-Focused Traditional Brand Identity Design Process
• 3. AI Logo Design Platform Process Design
• 3-1 Current Status of AI Logo Design Platforms
• 3-2 Analysis of The Process Structure & Design Creation Principles of
• AI Logo Design Platforms
• 3-3 Comparative Analysis of AI Logo Design Platform & The Traditional Process
• 3-4 User Research
• 3-4-1 User Research Design
• 3-4-2 First Prototyping Production・Pilot Test
• 3-4-3 Pilot Test Results
• 3-4-4 Secondary Prototyping Production・User Type-Specific
• Usage & Perception Survey
• 3-4-5 Analysis of Survey Results
• 3-5 A Proposal for an Efficient Design Process for The AI Logo Design Platform
• 4. Generative AI Learning Process Design for Brand Identity Design
• 4-1 Understanding Generative AI
• 4-1-1 Current Status of Generative AI Utilization
• 4-1-2 Prompt Engineering
• 4-1-3 Generative AI Functions & Research on Learning Methods
• 4-2 Exploring Brand Logo Design Using Generative AI
• 4-2-1 Brand Logo Design Experiments & Analysis of Design Generation
• Principles Using AI DALL-E
• 4-2-2 Establishing Generative AI Learning Method
• 4-2-3 Generative AI Learning Process Design
• 4-3 User Research
• 4-3-1 User Research Design
• 4-3-2 Pilot Test
• 4-3-3 In-depth Test of Usage & Perception by User Type
• 4-3-4 User Survey
• 4-3-5 Analysis of Survey Results
• 4-4 Exploring Brand Application Design Using AI Learning Process
• 4-4-1 Application Items Design Experiments
• 4-4-2 Derivation of Collaboration Factors & Methods
• 4-4-3 Detailed Division of Labor Between Designers & AI
• 4-5 A Proposal for an Efficient Generative AI Learning Process for
• Brand Identity Design
• 4-6 Comprehensive Comparative Analysis of AI Logo Design Platform Process
• & Generative AI Learning Process
• 5. Case Research Applying The Process
• 5-1 Selection of the Process Application Case・Research Purpose & Methods
• 5-2 Prism Energy Rebranding
• 5-3 Rebranding Design
• 5-3-1 Rebranding Design Process
• 5-3-2 STRATEGY
• 5-3-3 BASIC
• 5-3-4 APPLICATION
• 5-4 Design Results Awareness Survey
• 5-5 Case Research Results & The Creative Effects of AI Utilization
• 6. Conclusion
• 6-1 Research Results
• 6-2 Suggestions
• 6-3 Limitations of The Research
• References
• Abstract in Korean