Collaboration:
My master thesis project in collaboration with Toyota Material Handling Europe (Mjölby, Sweden)
My role:
UX designer and Researcher
Timeline:
5 months (2022-2023)
Background:
During this project, I explored the intersection of Artificial Intelligence (AI) and User Experience (UX) design, aiming to leverage AI capabilities to enhance the overall user experience design process.
Design Process:
I used the Research through Design methodology for structuring my research and the Double-diamond design methodology for the UX design part of the project. The design process commenced with an extensive literature review focusing on AI integration in the design process and designers’ willingness to adopt AI tools. This was followed by user interviews, data collection, and an exploration of generative AI tools. Subsequently, the process involved facilitating workshops, prototyping ideas, and employing AI tools to assess their viability as a design assistant. The research findings underscore the significant capabilities of AI in improving workflow efficiency for UX designers at all stages of the design process.
Interviews
Interview 1
Interview 2
interview 3
Interview 4
Focus group
I initiated a focus group session to determine the typical design process employed in a UX Design project. The objective was to pinpoint specific stages within the design practice where generative AI could offer valuable contributions.
Following an exploration of the prevalent design processes favored by UX designers and an assessment of various AI tools in design, including AI-powered Figma plugins, I proposed a selection of tools suitable for each stage of the process. Subsequently, these tools were put to the test during a hypothetical design project in a workshop, involving active participation from the workshop participants.
Workshop
I conducted a workshop on integrating AI tools like Eilla and ChatGPT in design ideation. Participants shared traditional ideation experiences, used AI tools for idea generation, and conducted a comparative analysis. While AI facilitated speedy benchmarking, a limitation was noted: generated ideas often mirrored existing concepts due to reliance on pre-existing data. The workshop concluded with participants gaining a nuanced understanding of when to use AI tools versus traditional methods, empowering them to make informed decisions about AI integration for optimal design outcomes.
Persona
By leveraging insights gathered from interviews and workshops, I developed a Persona, enabling me to center my focus on the target user group. Crafting a Persona based on insights gathered from interviews and workshops proved instrumental in refining our project’s target audience. To facilitate this, I utilized Fabrie, an AI tool, for persona creation and brainstorming. Fabrie offers a range of templates and a collaborative online workspace, which allowed me to experiment with the tool’s capabilities and assess its strengths and limitations from a designer’s perspective.
Brainstorming
Low-fidelity Prototype
I was eager to explore the feasibility of my initial concepts by creating a rapid prototype with Chat-Gpt. To achieve this, I requested the code, made minor adjustments, and then integrated it into JSFiddle to generate a visual representation of the code. The images below depict the preliminary low-fidelity prototypes.
Hand-drawn wireframing
Following this, I proceeded to draft the wireframe for the AIUX app, designed to serve as an assistant for UX designers. This application offers designers a streamlined platform to input project data by responding to a series of quick questions. Subsequently, the app generates a board where AI examines the provided data, as well as data from the internet, and offers suggestions at each stage of the project. Designers have the freedom to accept these recommendations as they see fit or ask for more information.
High-Fidelity prototype
Outcomes:
Based on this research, I recommended the development of a dedicated application that specifically addresses the needs and workflow of UX designers, seamlessly incorporating AI capabilities throughout the design process. By aligning this application with the needs of UX designers, it would create a more seamless and efficient experience, ultimately enhancing their productivity and allowing them to harness AI more effectively.
Usability testing and evaluation
To assess the Figma prototype’s usability, I organized a usability test involving design students. Additionally, I shared the Figma prototype URL with two expert UX designers for their evaluation. In the usability test, participants engaged with the prototype from start to finish and subsequently shared their valuable feedback and insights. I recorded their comments and suggestions for future consideration.
Reflection:
This project has been a remarkable learning journey for me. I’ve deepened my understanding of machine learning. I’ve also explored the vast capabilities and inherent limitations of AI, which have broadened my perspective. Furthermore, this project has allowed me to refine my skills in using Figma and app design.