Cosmo

Degree Project 2024

My thesis project explores how AI can inspire and support industrial designers throughout the form exploration process. Using a human-centered design approach, the project led to the development of COSMO, an AI-integrated, inspiration-triggering mood board tool for industrial designers. It addresses the issue of limited variety in current digital inspiration platforms, where similar content is often suggested due to AI algorithms. COSMO encourages using a variety of inspiration sources, including digital visuals and real-life experiences. This approach fosters creativity and personal connection, helping designers deeply resonate with the inspiration they collect and retrieve it at a more profound level. COSMO also utilizes Mixed Reality technology to offer an immersive, distraction-free experience when organizing and internalizing inspiration sources. It offers a meditative space that uses AI-assisted association loops. These loops create multiple layers of associations, assisting human designers in widening their initial associations. This leads to inspiration sources that are semantically different but insightful.

Project Information

A mood board is a type of visual presentation or 'collage' consisting of images, text, and samples of objects in a composition. It is widely used for inspiration during the process of form exploration in industrial design practices.

Nowadays, industrial designers use digital inspiration collection platforms such as Pinterest to navigate through the ocean of visual images in search of inspiration for form exploration, eventually clustering them into mood boards. However, they frequently encounter similar collections of inspiration due to AI algorithms suggesting similar content, which ends up cycling within the loop of closely related associations. This deviates from the essential principle of deriving inspiration, which involves creating indirect, distantly related associations to broaden creative boundaries and enhance the quality of inspiration. 

Furthermore, inspiration involves incorporating a variety of sources. While collecting visuals from digital platforms is one method, these flat images often lack a personal connection. I propose that an inspirational mood board should be a collection derived not only from digital visuals but also from individual life experiences. These scattered, fragmented, yet inspiring moments are what we, as humans, truly connect and resonate with. 

Methods

The project adopts a human-centered design approach. This method starts by identifying and understanding user needs, creating and prototyping potential solutions, testing with users, and refining the solutions based on their feedback

While I didn't strictly follow this process in a linear manner, often jumping between stages to gain user insights and frequently changing prototyping directions. The overall method comprises 5 main stages: research, empathise & define, prototype & ideation, user testing, refinement & implement.

Results

The result of my exploration is COSMO, an AI-integrated, inspiration-triggering mood board tool that enables industrial designers to collect inspiration sources about forms from different media, such as digital platforms and real-life events, and provide an immersive virtual meditative space for industrial designers to focus on building meaningful connections and making sense of those inspiration sources to turn them into mood boards that truly resonate and inspire during the form exploration process.

The COSMO system will inspire industrial designers in 3 scenarios

  1. Collecting inspiration sources
  2. Gathering & sorting inspiration sources
  3. Internalizing & making sense of inspiration collections

(1) Collecting inspiration sources

COSMO utilizes Mixed Reality (MR) technology to offer designers an immersive experience for sourcing inspiration from real-life events. Designers can capture these inspirations easily by pointing and drawing with their MR eyewear. The AI in the COSMO system then automatically categorizes and organizes these inputs into the Inspiration Hub. This hub serves as a collection gallery for all the gathered inspiration sources.

(2) Gathering & sorting inspiration sources

The Inspiration Hub's AI system sorts and categorizes all collected inspiration sources. This intelligent system allows designers to explore the inspiration sources in a physical space using MR technology. Designers can select specific sources that they want to explore further and move them from the physical space to the "Top of Mind" section in the Inspiration Hub. This action indicates to the AI system that the designers will work further on these sources, trying to understand and transform them into useful insights for form development in industrial design.

(3) Internalizing & making sense of inspiration collections

The COSMO system creates a virtual meditative space for designers to engage deeply with the inspirational thoughts collected in "Top of Mind." This space offers a mechanism for association loops. In this process, humans trigger the initial associations by selecting specific inspiration sources from "Top of Mind." The AI then suggests the first layer of directly relevant associations.

From these suggestions, humans choose and reflect on these associations verbally. The AI then summarizes these reflections into keywords to form a second layer of associations. These are less directly related to the initial associations. As more layers of associations are formed, the later ones deviate from the initial associations, resulting in inspiration sources that are semantically different yet insightful.

Designers maintain full autonomy in this process. They can stop the association loop once they find the current layer of associations insightful enough, allowing them to move to the next stage of ideation.

Hao Zheng

Master's Programme in Interaction Design
Hao Zheng Portrait
UID24 – Hao Zheng Grad Project
UID24 – Hao Zheng Grad Project
COSMO will help designers for inspiration all the way from collecting inspiration sources, gathering & sorting inspiration sources to internalising inspiration collections.
Idea quality in a meditative space is enhanced by numerous association loops with AI, leading to distant associations and high-quality ideas.

Idea quality in a meditative space is enhanced by numerous association loops with AI, leading to distant associations and high-quality ideas.