Creating an AI-powered Mechanical Engineering R&D System (MVP)

Timeline

Jan - Jun 2024

Platform

Desktop

Role

Lead Product Designer

Team

1x Product designers, 4x Software developers, 1x Algorithm engineer

Overview

GEAR is an AI-powered platform designed to revolutionize the engineering R&D process by enhancing efficiency and streamlining workflows. Initially launched as the world’s first AI-driven drawing checker, GEAR made a significant impact during its MVP stage, securing $500K in angel funding and attracting major clients like Volvo Truck and Eddie Precision Machinery. Since then, the platform has expanded its functionality and iterated to version 1.2, continuing to drive innovation and deliver value to its users.
My Role
As the design lead I contributed on both strategic and hands-on levels:
  • Provided leadership to a team of two designers.

  • Created a holistic design vision that guided our product strategy.

  • Collaborated with a cross-functional team to design and deliver the product.

  • Successfully led beta testing with Volvo, achieving product validation.

  • Co-presented venture capital pitches, securing angel funding.

Overall progress

Jan-Apr 2024 (MVP)
AI Based Drawing Checker

The AI-driven system enables 2D drawing review, automatic dimension and tolerance ballooning, GD&T data extraction, incorrect GD&T recognition, and facilitates online collaboration between QA and design engineers.

Apr-Jul 2024 (V1.0)
AI Based Engineering Assistance

The AI-powered system provides engineering suggestions aligned with ASME/ISO standards and material handbooks, breaks down requirements into design component specifications and engineering needs.

Aug-Nov 2024 (V1.1)
AI Based CAD Design Support

Our AI Based Automated 3D Model Generation can generate corresponding 3D models based on input design requirements to expedite the design process, helping design engineer to brainstorming better concept designs.

MVP

MVP Stage Business Goal
1. Define Key Features for MVP

Focus on identifying and prioritizing the essential features that deliver maximum value to users while meeting core business objectives, ensuring a streamlined and impactful first release.

2. Acquire Industry Leaders as Early Adopters

Partner with leading organizations to validate the product, build credibility, and generate case studies for broader adoption.

3. Secure Funding

Attract investment to enhance AI capabilities and introduce advanced features, ensuring the platform remains innovative and competitive.

MVP Stage Key Metrics
$
5
4
2
0
1
3
5
5
5
0
0
6
4
2
0
1
3
0
0
0
0
6
4
2
0
1
3
0
0
K

Angel Funding

1
4
2
0
1
3
5
5
0
0
0
6
4
2
0
1
3
0
0
0
0
6
4
2
0
1
3
0
0
%

Target Customers Adoption

1
4
2
0
1
3
5
5
0
0
0
6
4
2
0
1
3
0
0
0
0
6
4
2
0
1
3
0
0
%

Beta Test Task Completion

Shape from ambiguity

Research activites
  • Stakeholder Interviews

  • User Research Surveys

  • User Interviews

  • Competitor Analysis

Key insights from user research
How was mechanical design engineers previous experience during mechanical R&D process?
1. The CAD drawing review process is becoming a bottleneck

The CAD drawing review process is becoming a bottleneck, slowing down the overall workflow.

2. Disjointed workflow

There are 7 user roles and 11 steps involved in the mechanical R&D process. Various software tools are utilized, leading to numerous file transfers.

3. Poor user experience

Tools are hard to understand due to complexed workflow and the layout is outdated.

Key insights from competitor analysis
1. 100% AI adoption rate

All three competitors leveraged early stage AI technologies in different aspect.

2. Key features: workflow optimization, AI-driven insights

Competitors highlight features are workflow optimization, or AI-driven insights.

3. Monitization: B2B

The user segments competitors are focusing on is enterprises.

Vision & Mission

Vision
To revolutionize the mechanical R&D process by empowering them with intelligent, automated solutions.
Mission
To help mechanical R&D automate repetitive tasks, make smarter decisions, and improve operational efficiency.

Business Canvas

Process

Final design

As a drawing checker, I want to review the checking result of a uploaded drawing.

John Staff

Drawing checker

Design Challenge

As a drawing checker, I find the AI checking result is wrong, and I would like to manually edit it and submit the error type made by AI.

John Staff

Drawing checker

Building the human-centered AI tool

Reflections
1. AI feedback loop for intuitive mechanisms

Designing intuitive mechanisms for users to provide feedback on AI outputs required balancing simplicity with depth—ensuring it was easy for users to engage without interrupting their workflow.

2. Transparent communication with stakeholders

Gaining trust with stakeholders is crucial for ensuring product success, especially when navigating strategic decisions and iterations.

3. Success through aligning client expectations

Presenting a roadmap of opportunities, including how their input could directly influence the platform's evolution, transformed the conversation from one of transactional testing to a partnership-driven collaboration.