Jabil Manufacturing Dashboards

Summary: The importance of validating a project with users before writing code.

Overview

The Problem

Jabil Innovation Lab was looking to provide more insight and transparency to its supply chain partner operations, in particular, the distributed manufacturing sector. Across the globe, Jabil helps to manage and structure manufacturing processes for numerous industries.

Jabil provided us with a 45page SoW put together by the project stakeholder through numerous line managers and other key stakeholders. The goal was to deliver a single dashboard with over 35 data points that could serve as a ‘foundation’ across Jabil’s global manufacturing facilities.

Solution

It was clear early on, that the original SoW was only going to result in some very unhappy principles. Any dashboard, in order to be even mildly useful, HAS to provide it’s intended audience with some sort of actionable data. In order to do this, we also need to understand the environment the dashboard is intended to be displayed along with environmental constraints. Our solution was to instead deliver 3 unique dashboards that could serve the key roles located across the various manufacturing facilities.

Our Approach

Research

  • Task Analysis
  • Interviews (Stakeholders, Productions Managers)
  • Contextual Inquiries
  • Surveys
  • Proto-Personas
  • Journey Maps

Design

  • Informed Brainstorming
  • Sketching
  • Remote-Live Design Sessions
  • User Feedback
  • Cognitive Walkthroughs
  • Wireframes

Assessment

  • Cognitive Walkthroughs
  • Questionnaires

My Contribution

During our discovery phase, I began with interviews of the Projects key stakeholder and author of the SoW. After those initial interviews, I conducted a series of short interviews with some of the other contributors to the SoW to identify and clarify their intended ‘objectives’ and how they were going to define the ‘success’ of the project. It was during these interviews that I also was able to get clarity on the environment these primary dashboards were going to be displayed. 

I followed up on these interviews with a number of ‘Remote-Live Design Sessions”. During these sessions, I would work with my Project Manager as a facilitator, while I shared my screen with a handful of participants. We would then collectively design each dashboard according to the requirements necessary at that particular plant. This was repeated for a dozen or so other facilities until it was believed we had a significant sample size to determine a solid median of data.

Yield Dashboard Wireframe
Defects Dashboard Wireframe
Production Dashboard Wireframe

I synthesized the results of these ‘Remote-live sessions’ and developed a collection of 3 aggregate dashboard wireframes for the key roles that had been identified. These dashboards each then went through two rounds of moderated cognitive walkthrough and usability sessions that I lead.

Research

Surveys & Interviews

The interview sessions consisted of small groups of 5 from around the world. Each group reflected either assembly line workers or the managers, along with my project manager as a facilitator.

I pre-wired a blank browser window set to the final output screen resolution of 1024×768. This sounds like a lot of screen real estate until you start to populate the screen, keeping in mind that the end users had no means of interacting with the dashboards, let alone, scrolling or clicking.

Each group was asked 2 questions:

What is the one piece of information a line worker needs to know they are meeting the target yield?

What is the one piece of information a line worker needs to know to correct or improve meeting the target yield?

Key Takeaways

Based on the survey and interview results, we realized early on that a single dashboard was not going to work for the key targeted user groups that we were able to define.

Since we were able to narrow down our users into three groups based on evidence gathered instead of relying on arbitrary anecdotal evidence from a group of stakeholders, we identified the primary needs of each possible group in the manufacturing process and developed the dashboards based on those needs.

Redefining Our Users

Consistent with our interviews and contextual inquiries, we decided to focus on developing individual dashboards for three key ‘roles’ within the manufacturing process.

From our interviews, we learned that the dashboards we would come to call the “Yield” and “Defects” dashboards would be displayed on television sets over the assembly lines. The primary users of these dashboards would have no input devices accessible and they would be limited by the overall speed of the line.

Proto Personas

As mentioned in the key takeaways, our interviews had identified three key user groups that would eventually be using these dashboards.

Below are a couple of the Proto-Personas developed for this project.  I have written an article published on Medium, Better Personas: One Simple Change to Help Mitigate Bias to explain why I believe  Proto-Personas are significantly more helpful.

Final Design

Our final design for the Jabil manufacturing dashboards resulted in the following screens to be used as templates across Jabil’s global manufacturing facilities.

These dashboards would be getting displayed on 32″ screens mounted near key positions on each assembly line. 

Yield Dashboard
Defects Dashboard

Production Dashboard iPad Application

Reflections

There is a number of things I could go into about this project.

Working with a global client…
With team members located across the globe…
Worse yet vaguely (and that’s being generous) defined users…
Within the constraints of a tight budget, an even tighter deadline…
With a hastily thrown together requirements document…

But, none of that matters.

At the end of the day, I am happy with the results we achieved. More importantly, our client was happy. As is often the nature in Enterprise UX, in particular from the consulting side, there is rarely any ‘follow-up’ with the actual users. In this case, we did hear back a few months after the project began rolling out across all of Jabil’s manufacturing facilities and the responses were overwhelmingly positive, especially from the line workers who now knew how well they were performing. 

That’s what matters, to me anyway. Yes, I’m sure that Jabil’s numbers positively moved and I’m sure production increased globally. Those are all great for Jabil and I am glad I could do my part, but I get more satisfaction of knowing those line workers, or line supervisors might have just a little bit better day thanks to a few dashboards I designed.