About the product

REFIT, developed by Northwestern University’s Center for Deep Learning and built by the Departments of Computer Science and Industrial Engineering and Management Sciences, offers assurance to industries, ranging from agriculture to health care to manufacturing. REFIT’s system is built to consume and capitalize on IoT infrastructure. Using machine learning with state-of-the-art artificial intelligence (AI), REFIT’s predictive results and ease of use allow users to quickly develop, experiment, and deploy a robust, powerful, feature-rich IoT framework, making work more efficient with the same resources. The design allows it to evolve and continuously improve its predictive abilities, while not disrupting the deployment of the current machine-learning or deep-learning approaches

Digital Product

4 months

Duration

Product Type

My Role

Product and Project Manager

Tools Used

Empathize

Understanding People

Step 1 - Define the Research Method

Before conducting the research and user interviews, it is recommended to have a plan in place. A research plan is a document you create to list your research goals, as well as the methods you intend to use to accomplish those goals. It will also have an estimated timeline of the entire process.

Step 2 - Market Size

B2B Professionals in Manufacturing:

• Product Engineers: Enhancing efficiency in documentation & supporting impactful design decisions

• Project Managers: Simplifying oversight, streamlining approval processes, and ensuring timely & successful product launches

B2C and Educational Institutions:

• Students and Enthusiasts: Individuals seeking an accessible tool for honing their design skills without the complexity of traditional CAD software.

Step 3 - Comparing it with existing Solutions

Currently, our target customers employ various methods to execute their Engineering Documentation processes, but they face distinct challenges:

Market Gap

• Inadequate inventory management

• Quality control challenges

• Rising labor and operational expenses

• Unreliable product delivery schedules

• Vulnerability to supply chain disruptions (e.g., supplier problems and natural disasters)

• Limited adoption of innovative technologies to enhance efficiency

Step 4 - Primary Research (Surveys & Interviews)

After going through all the data in secondary research, it helped us to make some hypotheses but it was equally important to validate those. 

We interviewed individuals from different cohorts and occupations: Product Engineers, Design Engineers, Production Engineers, and Maintenance Engineers.

These interviews helped with qualitative data and found the user's needs and concerns.

Define

Figuring out the problem

Step 2 - What are all possible markets?

Synthesizing the user interviews and survey results, I did an affinity mapping, where I tried putting all the similar experiences together. This gave a good idea about the number of users facing similar challenges and a couple of good insights about the users and their challenges.

Step 1 - Choosing the right metrics

AARRR Framework - AARRR stands for Acquisition, Activation, Retention, Referral & Revenue. These are the 5 most important metrics for the growth of any company.

Activation: Time Efficiency can be tied to the Activation stage, where the focus is on the user’s initial experience with the product. If SelfCAD helps users quickly and efficiently complete tasks, it enhances their first impression, leading to a higher likelihood of continued use and satisfaction.

Retention: Retention Rate directly aligns with the Retention stage. It tracks the percentage of returning users, indicating how well your product engages users over time and keeps them coming back. A strong retention rate is crucial for sustainable growth and reducing churn.

Referral (NPS): The Net Promoter Score (NPS) is an important metric within the Referral stage. It measures customer satisfaction and their likelihood of recommending your product to others. A high NPS suggests that users are satisfied and could become advocates, driving organic growth through word-of-mouth.

Revenue: Conversion Rate is directly tied to the Revenue stage. It measures the percentage of users who upgrade from free to premium plans, reflecting the success of monetizing your user base. High conversion rates indicate that the product offers significant value that users are willing to pay for.

Step 3 - Prioritized Target Market

• Mid-sized discrete manufacturing companies

• Although companies have invested in connected machines and sensors they struggle capitalizing on the data.

• Manufacturing prioritizes cost, quality, and reliability and supports economic recovery, and increased demand for American goods.

• REFIT strengths include cost saving and focus on reliability and quality

Step 4 - Interview Insights

The pervasive issue of downtime results in an average loss of production hours annually. The reliance on manual, retrospective data collection methods obstructs operational insights resulting in productivity loss and loss of competitive advantage. Solving this challenge is imperative for companies seeking to remain competitive in an evolving and data-driven industry landscape.

Ideate

Generating your ideas

Step 1 - How might we(HMW)?

How Might We (HMW) questions are one of the best ways to open up Ideation sessions. By framing the problem as a How Might We question, we will prepare ourselves for an innovative solution.

• How might we automate the generation of standard design documentation?

• How might we facilitate seamless collaboration between designers, engineers, and other stakeholders?

• How might we provide real-time collaboration and feedback mechanisms to speed up the documentation workflow?

• How might we cater to the specific requirements of small-scale companies and startups?

Prototype

Creation and experimentation

Step 1 - Sketching - Pencil before Pixel

I usually start the design process with low-fidelity wireframes. This is the way I iterate through many design options quickly.

Step 2 - Low Fidelity Wireframes

While designing the wireframes, there were a couple of feedback loops from the users to ensure that I was moving in the right direction.

Step 3 - Design Principles used for SelfCAD

1. Effortless Navigation: Design for intuitive access and minimal complexity, enhancing user efficiency.

2. Visual Clarity: Use clear visual elements for faster, more effective user comprehension.

3. Streamlined Workflows: Simplify common tasks to boost productivity and save time.

4. Engaging Experience: Ensure the software is interactive and enjoyable to encourage user satisfaction.

5. User-Centric Design: Focus on users' needs and challenges to ensure practical utility and relevance.

6. Adaptive Interface: Cater to diverse user preferences with a flexible and customizable design.

Step 4 - High-Fidelity Wireframes

• Detail Part Previews: After uploading the documents, engineers can check the files that have been uploaded and remove or add files as required. Moreover, they can measure parts, view individual parts, and view sections of assembly or parts.

• Edit Screen: On the edit screen, the product engineers can collaborate and work on a document simultaneously and make changes as recommended by the quality assurance team who can mark and leave comments, and assign a list of to-do list style work completion.

• Approvals Screen 2: The page is an extended version of changes that shows approvals from other stakeholders. This provides the following process of getting approvals from all the stakeholders at the same place.

Final Result

Customer Centricity - I could have never thought of improving the designing and documentation process making the app more user-friendly.

Usability Testing - This was a completely new experience for me. Every user has their own journey and ways of using the app. It is very important to understand the goal and mindset of the user and what they are seeking at the point.

3 months

15+

Screens

Timeline

10+

Iterations