teammates
Yaqi Wei (Team leader)
Jiani Wang
Qijun Nie
My role
User research
Visual design
Concept
User flow
Wireframe
Prototype
tools
Figma
Procreate
Photoshop
MG
Notion
Slack
timeline
0-3 week: Research & Concept
4-6 week: Iteration & Testing
7-9 week: Prototype & Testing
Creating a seamless workflow for architects: An AI-powered work delivery and management platform
As an ex-architect, I have found that UK architects and managers, despite their immense workloads, find it difficult to derive individual growth and a sense of value recognition from their work, and that they have a weaker connection to projects and each other. Against the backdrop of global digitalisation, the overly bloated traditional management systems of the construction industry have greatly impacted on people's productivityTo boost productivity and fulfillment at work, we want to build a flexible platform that brings teams and managers together flexibly based on project demands, simplifies useless processes, and cares for people's personal growth.
Emphasis
Interviews, survey, group interviews
Before we started our research, we had some initial assumptions about the problems with construction management systems based on our second-hand research into the construction industry and my personal internship experience. Before the interview, I prepared the interview script and had my teammates review it, and after three revisions I finally completed the first interview script.
In the interviews with architects and managers, we have found too many pain points to narrow down.
To narrow down the pain points, we held the group interview and drew a user journey map together, which help us to align the goals. In the end of the interview, the participants voted for the most important pain points.
Analysis and define
Personas, empathy map
I analysed the information I gathered and created user personas and empathy maps. Defining users and scenarios helped me understand who uses the product and what problems they want to solve.
Ideate and design
Wireframe, prototype, usability-testing
By demonstrating how AI can analyze vast amounts of data and produce real-time insights in natural language, ChatGPT has transformed the way people perceive the capabilities of AI. However, AI could be conveying wrong messages and even poses a threat to human behaviour. Thus, we did research on how to design-powered system safely and effectively.How AI is applied is the focus of the project, and based on the likely needs of architects and managers, and in conjunction with Microsoft's Human AI Interaction Guideline, we have initially identified four ways in which AI can be applied. Among them, I focused on the parts of AI precheck and AI chatbox, paying attention to the seamlessness of human-AI interaction in the design.
Outcome and takeaway
Prototype, design system
This project has been a very rewarding journey for me, combining my skills as a product designer in user research and product design with my ability as an architect to build product architecture and a high aesthetic. Most importantly, as a technology enthusiast, I managed to speak to people in the architectural community about the potential of AI applications as a solution, and received positive feedback and support from them. In the design process, we maintain human-centred values, think dialectically about the pros and cons of AI in practical applications, and put the needs and behaviour of architects and managers at the forefront.
solutions
Smart task management
We work to enable the AI to learn from past company data and help managers to break down a new project brief into stages containing many tasks as required by RIBA, inserting them into existing timelines and suggesting assignments to the right team members.
Smart collaboration
We use AI in synergy with people management to streamline the approval process for the use of resources by architects; in cross-departmental collaboration, AI is responsible for monitoring the progress of team members' tasks and making recommendations to managers for the coordination of task progress.
Smart project assistant
The AI schedules architects with multiple tasks by learning from the company's past cases and provides suitable reference cases as the architects complete their tasks; the key to improving the efficiency of the entire chain is the AI's precheck mechanism, allowing the architects to retain their autonomy while meeting the metrics required by their managers.
