UX Design - Final Compilation & Reflection
22/09/2025 - ( Week 1 - Week 14 )
Ho Winnie / 0364866
UX Design / Bachelor's of Design Honors In Creative Media
Final Compilation & Reflection
1. Task 1 - UX Study Exercise
Students are required to select an existing product or service and conduct a user experience (UX) analysis by identifying its key strengths and weaknesses. Based on the identified UX issues, students must propose practical design improvements that enhance usability and overall user satisfaction.
The study must be supported by a User Journey Map that visualizes user actions, emotions, pain points, and opportunities for improvement across the interaction stages.
Final Output :
Students must select one existing service or product and perform a full UX design process, beginning with user research and ending with a high-fidelity prototype. The project must include user interviews, affinity mapping to synthesize insights, persona creation, and problem definition grounded in real user data.
Based on the findings, students are required to propose and design UX improvements, supported by user flows and interaction logic, culminating in a polished high-fidelity output that demonstrates clear usability and design rationale.
Progression -
Our team consists of five members — Melvin Yung (Team Lead), Ho Winnie, Lin Si Yan, Lew Guo Ying, and Ng Kar Yee — with four of us specialising in UI/UX design. Leveraging our collective design expertise, we decided to pursue a project that extends beyond conventional mobile or web interfaces and explores an immersive, interaction-driven solution.
We selected AR Cook, an augmented reality cooking assistance service designed specifically for culinary students. Culinary education requires students to master complex techniques, precise timing, and strict safety procedures, often within fast-paced and high-pressure kitchen environments. Traditional learning methods such as printed recipes, static videos, or instructor demonstrations can be difficult to reference in real time while cooking.
AR Cook aims to support culinary students by overlaying step-by-step cooking guidance, visual cues, and safety reminders directly into their physical cooking space. By providing hands-free, contextual assistance, the service seeks to reduce cognitive load, minimize errors, and improve learning efficiency without interrupting the cooking process.
Through this project, our team aims to investigate how augmented reality can enhance experiential learning in culinary education, while applying UX principles to ensure that the AR experience remains intuitive, supportive, and non-intrusive for student chefs.
To ground our design decisions in existing practices, the team conducted a contextual study focusing on both AR cooking systems and AR hardware technologies. The study was divided to ensure depth across user experience, interaction design, and technical feasibility.
The contextual study on existing AR cooking systems was conducted by Me, Lew Guo Ying, and Lin Si Yan. This segment focused on analyzing how augmented reality is currently used to support cooking tasks, particularly in instructional and learning contexts.
The study examined commonly implemented features such as:
-
Step-by-step AR recipe overlays
-
Visual indicators for ingredients, measurements, and timing
-
Hands-free interaction methods (gesture and voice control)
-
Safety cues and task progression prompts
Notable existing AR cooking systems and prototypes include
-
Microsoft HoloLens – Kitchen Assistance Demonstrations
Microsoft has showcased multiple HoloLens demonstrations where step-by-step cooking instructions, timers, and visual cues are displayed directly within the user’s field of view while cooking. -
Panasonic AR Cooking Support System
Panasonic introduced an AR-based cooking support concept that projects recipe steps, ingredient guidance, and progress indicators onto kitchen surfaces to assist users during food preparation. -
Siemens Smart Kitchen AR Prototype
Siemens explored AR kitchen concepts where cooking instructions and appliance feedback (e.g. oven temperature and timers) are visually overlaid to support smart cooking workflows. -
HoloKitchen (Research Prototype)
HoloKitchen is an experimental AR cooking system developed for research purposes, focusing on projecting spatial instructions and procedural guidance during cooking tasks using head-mounted AR displays.
Strengths Identified
Across these systems, several UX strengths were consistently observed:
-
Context-aware instruction delivery, allowing users to receive guidance without shifting attention away from cooking
-
Hands-free access to information, particularly useful in environments where hands are occupied or unclean
-
Improved visual understanding of cooking techniques and procedural steps
-
Increased engagement through immersive, spatial interaction
These strengths highlight AR’s suitability for task-based and experiential learning contexts such as culinary training.
Key Weaknesses and Pain Points
Despite their innovation, the following UX limitations were identified across most existing systems:
-
Lack of personalisation
Most systems provide static, linear instructions that do not adapt to individual skill levels, learning pace, or user confidence. This is a significant limitation for culinary students, who progress at different speeds and require varying levels of guidance. -
Absence of crisis management and recovery support
Existing AR cooking systems largely assume ideal cooking conditions. Scenarios such as overcooking, missed steps, incorrect ingredient usage, or safety incidents (e.g. oil overheating) are rarely addressed. When errors occur, users are not guided on how to recover or adjust. -
Interface clutter and visual overload
Some implementations display excessive overlays, which can obstruct the cooking workspace and increase cognitive load, particularly in fast-paced kitchen environments. -
Limited educational focus
Most systems are designed for home users or technology demonstrations rather than structured culinary education, resulting in insufficient pedagogical scaffolding and feedback mechanisms.
These weaknesses reveal a clear gap between existing AR cooking technologies and the practical needs of culinary students. Attached below is the document I did when carrying out contextual studies.
Following the contextual study, the next phase of our UX process focused on primary user research to better understand the needs, challenges, and behaviours of our target users — culinary students. This step was critical in grounding our design decisions in real user experiences rather than assumptions.
To ensure both breadth and depth in data collection, we adopted a mixed-method approach consisting of interviews and a Google Form questionnaire.
The target audience for this study is culinary students, who regularly engage in hands-on cooking within training kitchens. These users often operate under time pressure, must follow precise procedures, and need to manage safety risks while learning new techniques.
To streamline the research process, responsibilities were divided as follows:
-
Questionnaire (Google Form): In which I will document the process
Designed and managed by Me and Ng Kar Yee -
User Interviews:
Conducted by Melvin Yung, Lin Si Yan, and Lew Guo Ying
This division allowed the team to collect both quantitative insights (patterns and trends) and qualitative insights (in-depth user experiences).
The Google Form was created to gather broad insights from culinary students regarding:
-
Their current learning methods during cooking
-
Pain points when following recipes or instructions in real time
-
Common mistakes or stressful situations encountered while cooking
-
Familiarity with and perceptions of AR or smart cooking tools
The questionnaire helped identify recurring challenges and validated whether the problems observed in our contextual study were also experienced by real users.
The survey results show that most culinary students practise cooking frequently and identify as beginner to intermediate learners, yet face significant challenges during real-time execution such as remembering steps, visualising techniques, controlling temperature, and managing time under pressure. One-on-one instructor feedback is limited, contributing to high stress levels during practical sessions and assessments.
Students are receptive to AR and AI assistance, particularly features that offer step-by-step visual guidance, real-time safety alerts, timer tracking, and personalised feedback. Crucially, the findings highlight a strong need for customisable guidance and crisis management, where the system can adapt to skill level and support users in recovering from mistakes rather than only guiding ideal cooking flows.
Following the analysis of our interview findings and Google Form responses, we moved into the affinity mapping stage to synthesise raw research data into meaningful insights. The purpose of this phase was to identify recurring patterns, shared pain points, and emerging user needs across our target audience of culinary students.
We initially began this process by creating a structured affinity map, where insights were pre-organised into defined categories such as technique, safety, timing, and learning support. While this approach allowed us to summarise findings efficiently, we were later advised that it limited discovery and risked reinforcing assumptions too early in the synthesis process.
Based on feedback, we shifted to a more exploratory approach using handwritten scratch notes and unstructured groupings. Individual observations, quotes, and behaviours from interviews and survey responses were written onto separate notes and then grouped organically based on natural relationships.
This bottom-up method allowed themes to emerge naturally, rather than being imposed upfront. Through iterative clustering and discussion, we began to identify stronger connections between issues such as stress, lack of feedback, and difficulty visualising techniques.
Key groupings that emerged included:
-
Human skill issues (difficulty visualising techniques, knife skills, accuracy)
-
Lack of feedback and communication barriers (hesitation to ask instructors, limited real-time guidance)
-
Stress and emotional pressure (timed assessments, fear of failure, discouragement)
-
Safety issues (cuts, burns, unsafe kitchen handling)
-
Instructional gaps (unclear demonstrations, recipe confusion, insufficient educator support)
-
Environmental constraints (limited kitchen space, budget, equipment)
This grouping process revealed that many problems are interconnected, with stress, safety, and skill-related issues often occurring simultaneously during cooking sessions.
Based on the insights derived from affinity mapping, the team translated key pain points and opportunities into How Might We (HMW) questions. This step helped reframe user frustrations into actionable, design-oriented problem statements, guiding the ideation phase without prematurely locking into solutions.
Each team member contributed HMW questions from their perspective, ensuring a diverse yet complementary range of problem framings grounded in our research findings.
Collectively, the HMW questions revealed several recurring themes:
-
The need for real-time reassurance and feedback
-
Support for error recovery and crisis management
-
Reduction of stress, fear, and cognitive overload
-
Improved safety awareness and confidence
-
Clear guidance that adapts to different skill levels
These questions served as a bridge between research synthesis and ideation, ensuring that the subsequent design concepts for AR Cook directly addressed real user needs rather than assumed problems.
In the next phase, we use these HMW questions to guide ideation and concept development, narrowing down which challenges AR Cook will prioritise in its UX solution.
With clear problem framings established through the How Might We (HMW) questions, the project progressed into defining the broader ecosystem and core user experiences. This phase focused on understanding who is involved, who we are designing for, and how users interact with AR Cook over time.
To ensure clarity and efficiency, responsibilities were divided based on team strengths.
The stakeholder map was developed by Ng Kar Yee and Melvin Yung to identify all parties involved in or affected by the AR Cook ecosystem. This included direct users, indirect users, institutional stakeholders, and technical stakeholders.
The stakeholder mapping process helped clarify:
-
Primary and secondary stakeholders
-
Relationships between users, educators, and institutions
-
Influence, interest, and responsibility across the system
By mapping these relationships, the team ensured that AR Cook’s design decisions considered not only end users, but also operational, educational, and safety-related stakeholders within culinary training environments.
Based on insights gathered from interviews, Google Form responses, and affinity mapping, we identified three key persona types that represent distinct user groups within the AR Cook ecosystem. These personas reflect differences in skill level, learning needs, and expectations, allowing us to design a system that adapts across experience levels rather than adopting a one-size-fits-all approach.
Persona 1: Beginner Culinary Student
Beginner users are typically new to formal culinary training and require high levels of guidance and reassurance. They often struggle with basic techniques such as knife handling, temperature control, and multitasking, and experience heightened stress during practical sessions. This group benefits most from step-by-step instructions, visual technique demonstrations, safety alerts, and confirmation feedback that helps build confidence.
Persona 2: Intermediate Culinary Student
Intermediate users possess foundational cooking skills but still face challenges when performing under time pressure or learning more complex techniques. They seek flexible guidance, preferring key highlights, reminders, and real-time feedback rather than full instructions. For this persona, AR Cook must support error recovery, timing management, and consistency, enabling skill refinement without overwhelming the user.
Persona 3: Professional / Advanced Cook
Professional or advanced users are confident in core techniques and workflows, but value tools that enhance efficiency, precision, and performance. Rather than instructional guidance, this persona benefits from minimal overlays, performance analytics, plating refinement, and optional feedback on technique optimisation. AR Cook for this group acts as an assistive performance tool rather than a teaching system.
The user journey mapping for AR Cook was fully developed by me synthesising research insights, personas, and HMW questions into a holistic view of the cooking experience across different skill levels. The journey maps were created for three personas — Beginner, Intermediate (Amateur), and Professional (Expert) — to capture how needs, emotions, and pain points evolve with experience on each step of cooking.
With a clear understanding of user needs established through research, affinity mapping, and How Might We (HMW) questions, the project moved into the ideation phase. This stage focused on translating abstract problem statements into tangible feature ideas for AR Cook.
The team collectively reviewed all HMW questions and selected a subset that best represented the most critical user pain points — particularly those related to confidence-building, real-time feedback, safety, timing, and plating support.
Following the ideation phase, the team moved into feature prioritisation to identify which concepts should be carried forward into the final design. This step focused on selecting common features that were consistently rated as valuable, feasible, and impactful across personas and HMW questions.
Rather than pursuing all ideas, we evaluated features based on:
-
Frequency of appearance across ideation clusters
-
Alignment with key user pain points (confidence, stress, safety, clarity)
-
Relevance across beginner, intermediate, and professional personas
-
Suitability for real-time AR interaction in a kitchen environment
After identifying the common key features, the team applied the MoSCoW prioritisation method to systematically evaluate and scope features for the final AR Cook solution. This framework helped us distinguish between essential functionalities and nice-to-have enhancements, ensuring that the final design remained focused, feasible, and aligned with user needs.
The MoSCoW method categorises features into four groups: Must Have, Should Have, Could Have, and Won’t Have.
After applying the MoSCoW prioritisation, the team further refined the shortlisted Must Have and Should Have features using a 2×2 Impact–Effort Matrix. This step helped us evaluate which features would deliver the highest user value while remaining realistic to implement within the project scope.
Core priorities centred on safety, real-time execution guidance, feedback and reassurance, time management, and adaptability across skill levels, ensuring that AR Cook directly addresses the most critical pain points faced by culinary students.
6.5 Proposed Main Features
The final ideation resulted in a cohesive set of AR Cook core features designed to support culinary students across different skill levels.
AR Cook begins with a mascot-led welcome experience that adapts its tone based on the user’s skill level (Beginner, Amateur, or Expert). A short guided tour introduces core AR controls such as gestures, voice commands, interface resizing, and safety cues, followed by system calibration including workspace scanning, hand tracking, and utensil recognition. Users can customise their experience by selecting a mascot and preferred guidance mode, allowing the system to match their learning comfort level.
During cooking, AR Cook provides interactive recipe guidance with step-by-step visual overlays, AI-generated remarks highlighting precautionary steps, and instructor demonstration videos that can be resized, repositioned, and controlled in speed. Ingredient preparation is supported through AI freshness checks, while knife practice adapts to skill level—using cutting grids for beginners and portion-consistency guidance for amateur and expert users.
To ensure safe and effective execution, AR Cook integrates real-time safety alerts for heat, hand positioning, and restricted touch points, delivered through visual signage, mascot prompts, and sound cues. Cooking execution is supported by replayable overlays, timelines, and progress indicators, helping users manage boiling, frying, and baking tasks while reducing cognitive load during multitasking.
Plating assistance is tailored by skill level, offering structured templates for beginners and instant visual feedback for more advanced users. At the end of each session, AR Cook provides adaptive AI feedback and ratings—encouraging and detailed for beginners, neutral and selective for amateurs, and direct and professional for experts. Performance is summarised through star ratings, feedback messages, and learning recommendations, reinforcing progress and motivating continued practice.
To communicate the concept, functionality, and user experience of AR Cook, the project concludes with a set of design-led outputs that demonstrate both the system’s usability and its real-world application within culinary education.
Poster
- A visual summary that communicates the core problem, key UX insights, and main features of AR Cook. The poster is designed for academic presentations and exhibitions, allowing audiences to quickly understand the concept and its value in culinary education.
- An animated showcase illustrating the AR Cook experience in action, including onboarding, real-time cooking guidance, safety alerts, and feedback. This output helps visualise how AR interactions and overlays function within a real kitchen environment.
- The website serves as a central platform for explaining the concept, interactions, and rationale behind AR Cook.
To ensure efficient collaboration and clear ownership, the final outputs were divided among team members.
-
Poster Design
The poster design is handled by Me and Lin Si Yan, focusing on visually communicating the concept, UX insights, and key features of AR Cook in a clear and engaging manner. -
Website Design & Development
The project website is developed by Me and Lew Guo Ying, where I defined the art style, contents and hi fidelity where Guo Ying aided in low fi production and final auto layout. -
AR Cook Animation – UI & Mascot Design
All AR interface panels and mascot designs used in the animation are created by Me and Lin Si Yan, ensuring consistency in visual language, interaction clarity, and tone across the experience. -
Animation Production
The animation process is handled by Lew Guo Ying and Melvin Yung, focusing on motion, transitions, and the visual flow of the AR Cook experience. Ng Kar Yee and Melvin worked on storyboarding , initial editing of video flow. -
Sound Design
Sound effects and audio elements for the animation are produced by Ng Kar Yee, enhancing feedback, safety alerts, and overall immersion. Filming & Voice Actors
The filming of AR Cook animation is by Melvin with voice actors Me , Lew Guo Ying and Lin Si Yan.
During the early stage of the project, from the initial proposal submission up to the affinity mapping phase, our team encountered several challenges, particularly in how our ideas were communicated through the presentation slides. While the core concept of AR Cook was well-received, the slides lacked structural clarity, visual hierarchy, and narrative flow, making it difficult to clearly convey our research direction and UX rationale.
Recognising these issues, I took the initiative to rebrand and restructure the entire slide deck. This involved redefining the overall layout, reorganising content into clearer sections, and establishing a more consistent visual system that better reflected a UX case study format. The rebranding helped align the slides with our design intent, improved readability, and ensured that insights from research and affinity mapping were communicated more effectively.
In this phase of the project, Lin Si Yan and I focused on developing the visual identity and interaction components for AR Cook, specifically the mascot design and AR UI panels. To streamline the workflow and maintain design consistency, we divided responsibilities based on individual strengths.
Lin Si Yan led the design of the AR Cook mascot, establishing its visual personality, expressions, and tone to support different user skill levels. The mascot plays a key role in guiding users, delivering feedback, and conveying safety alerts in a friendly yet appropriate manner.
I was responsible for creating the design kit and overall AR panel style, defining the visual language, layout structure, colour usage, typography, and interaction patterns to ensure clarity and usability within an AR environment.
For the AR Cook animation video, responsibilities were further refined. I designed all AR panels related to the introduction flow, dashboard, ingredient detection, and knife-cutting guidance, focusing on clear information hierarchy and minimal visual obstruction. Lin Si Yan handled the panels for cooking execution, plating, and reflective feedback, ensuring continuity in interaction flow and emotional tone during the later stages of the experience.
This structured division of tasks allowed us to maintain visual coherence across the AR experience while efficiently progressing toward the final animated output.
My AR Panels by Winnie HoFor the promotional poster, Lin Si Yan and I collaborated closely to communicate the AR Cook concept in a clear and visually engaging manner. I initiated the design process by developing the initial layout, visual structure, image selection, and overall framework, ensuring that the poster effectively highlighted the core idea, key features, and value proposition of AR Cook.
Following this, Lin Si Yan further refined the poster by enhancing visual hierarchy, typography, spacing, and overall polish. Her refinements improved clarity and cohesion, resulting in a more professional and exhibition-ready output. This collaborative workflow allowed us to efficiently iterate on the design while maintaining a consistent visual direction aligned with the rest of the project
For the project website, I collaborated with Lew Guo Ying to translate the AR Cook case study into a clear and structured digital format. The process began with content drafting, where I outlined and prepared the written material that documented the UX process, research insights, and final outcomes.
Using this content, Lew Guo Ying first developed a low-fidelity (lofi) website layout, focusing on information structure and content flow. Building upon this foundation, I then reworked the layout into a high-fidelity design, refining visual hierarchy, spacing, and interaction details to align with the overall AR Cook design language. Throughout this stage, Lew Guo Ying supported the process by auto-layouting and organising content, ensuring consistency and responsiveness across the website.
This collaborative approach allowed us to efficiently progress from conceptual structure to a polished final output, resulting in a website that clearly communicates both the UX case study and the AR Cook concept.
The project concluded with a final presentation, where we showcased the completed AR Cook concept and its supporting outputs to our lecturer. During the presentation, we walked through the full UX process, from research and ideation to the final poster, animation, and website, clearly demonstrating how AR Cook addresses the challenges faced by culinary students during real-time cooking.
Overall, we received positive feedback on the quality, clarity, and coherence of our outputs, particularly the visual execution, interaction flow, and the application of UX principles within an augmented reality context. Our lecturer noted that the concept was well thought out and effectively communicated.
At the same time, constructive feedback was given regarding potential areas for further exploration. Specifically, our lecturer expressed interest in seeing how the system could better handle crisis situations (such as cooking errors or unexpected outcomes), how different profile-based UIs could be more distinctly visualised, and how the system supports multitasking in high-pressure cooking environments. These comments highlighted valuable opportunities for future iteration and refinement of AR Cook, reinforcing the importance of adaptability and resilience in real-world UX design.
Based on the feedback from the final presentation, I focused specifically on improving crisis handling within AR Cook, particularly for high-risk kitchen situations such as pan fires and self-inflicted cuts. These scenarios were identified as critical moments where users require immediate, clear, and non-panic-inducing guidance.
In the case of a pan catching on fire, AR Cook is designed to trigger an immediate visual and audio alert, shifting the interface into an emergency mode. Non-essential UI elements are hidden, and the system displays clear, step-by-step emergency instructions such as turning off the heat source, covering the pan with a lid, and avoiding the use of water. The mascot’s tone becomes calm and authoritative, ensuring that guidance is delivered quickly without overwhelming the user.
For situations where a user cuts themselves, AR Cook switches to a safety-first response by detecting abnormal hand movement or distress cues. The system prompts the user to pause the task and provides basic first-aid guidance, such as stopping bleeding, cleaning the wound, and seeking instructor assistance. Visual cues highlight safe zones and encourage the user to step away from sharp tools before resuming cooking.
By incorporating these crisis-handling scenarios, AR Cook moves beyond ideal cooking flows and demonstrates its ability to support users during unexpected and high-stress situations, reinforcing safety, confidence, and trust in real kitchen environments.
Final Presentation Slides :
Copy of UX Design Final Presentation by Winnie Ho3. Feedback
4. Reflection
Observation
Throughout the project, I observed that designing for AR requires a very different mindset compared to traditional UI/UX work. Factors such as safety, cognitive load, timing, and physical context play a much more critical role. I also noticed how important collaboration and clear task distribution were, especially when managing complex deliverables like animations, posters, and websites. Guidance from our lecturer was crucial in helping us refine our thinking, particularly in areas such as crisis handling and multitasking, which are often overlooked in early design stages.
Findings
From this project, I learned that AR cooking is far from simple, especially in real-world kitchen environments where mistakes, stress, and safety risks are common. However, with the right technological support, AR has strong potential to greatly assist culinary students by providing real-time guidance, reassurance, and safety support. This project reinforced the importance of designing beyond ideal user flows and considering edge cases and crisis scenarios, ultimately highlighting how thoughtful UX design can make complex tasks more approachable and effective.
Comments
Post a Comment