Problem
Large fleets of vehicles are hard to manage
After creating a bulk action to affect a vehicle fleet, owners need to approve further changes, monitor progress, and address any issues.
One of Rivian's clients, Amazon, ordered 100,000 electric vehicles for their Prime delivery service, presenting a challenge in fleet management.

Specifically, Amazon fleet owners required scalable solutions for managing bulk actions such as connectivity changes, eSIM swaps, and vehicle status updates across their expansive fleet.
Rivian's electric prime delivery vehicles
Solution
Consolidate real-time data and vehicle status
This dashboard includes key metrics and intuitive filtering, allowing fleet managers to quickly oversee vehicle statuses and focus on certain vehicles.
Empower filtering & searching
By helping users find vehicles based on attributes such as location or status, users can focus on vehicles needing their attention.
Filtering vehicle list based on location & status
Provide context into Who, What, Why
The Details tab gives high level information and settings for the bulk action. This view shows the reason description, cost analysis breakdown, and stakeholders involved.
Bulk Action's details, cost analysis, and stakeholders
Project Glossary
Fleet ~ group of vehicles under the same ownership.
Bulk Action ~ Complete a certain action on a group of vehicles at once.
(as opposed to completing the same action individually to each vehicle)
Connectivity ~ interconnection of platforms, systems, and applications.
(change cellular configurations, move carrier account, activate SIM card)
Research
Interviewing a wide range of user & roles
I interviewed current fleet owners to get a better understanding of their expectations of bulk actions and their current routines in managing their vehicles. I grouped users into three categories: watcher, approver, and creator:
Insights from users:
Checking in would not be very frequent. Customers will not check in often. Actions like profile swap and eSIM swap might be analyzed for a few days”
“First, I want to see the vehicles and the progress. I want know if we can retry this operation or roll back? ... it would be helpful to see error classification and then do any remediation possible for CMP(Connectivity Management Platform)”
INSIGHTS
Discovered insights are not actionable
Constantly waiting and checking on updates by revisiting and refreshing page
Easy to forget bulk action context, as most users do not regularly visit
Does not account for external users, overwhelmed or not enough context
Layout
Laying out feature priorities
Because watcher and approver user groups do not regularly check in, it was important to include giving context for the bulk action. On the other hand, the creator of the action needs updated information and allow them to directly troubleshoot vehicles that had problems completing the action.
To prioritize the key information needed by the users, I started by ideating on possible blocks or cards of information for the screen.
To address this, the static features are included to help provide context, while the non-static elements are updated over time to allow the user to troubleshoot.
Splitting Bulk Action dashboard into two tabs: Details and Progress
While it started with one page with all of the information to user everything at once, the plan shifted to a double page layout in order to direct the user’s focus to give context first and then show them progress that has been made to complete the bulk action.
As shown below, the layout of the progress and detail screens were iterated on further to get a sense of prioritization between various feature components.
Ideation
Creating a readable dashboard
Based on the user flows and the product requirement gathered, I designed multiple versions of each interface to diverge and weigh the pros and cons.
Going back the user pain points, users don’t have the time to click through all of their vehicles to get updates on a bulk action.
Weighing between filter designs
To address this, filters and progress visualizations were utilized to give a high level overview to track a bulk action.
Visualizing fleet progress
A data visualization was needed to show the progress of the action over time. Each vehicle’s status was determined by its stage of completing the action such as Scheduled, Error, and Completed.
The progress visualization shows the distribution of the vehicles’ statuses at a quick glance.
Initial exploration
I started brainstorming different directions of how to visualize the vehicle statuses. The most prominent options included the line graph, area chart, pie chart, and progress bar. Based on user and fellow design feedback, the bar was the most clear to understand at a quick glance.
Digging deeper...
I tweaked the progress bar so that the percentages for each status was still readable no matter how small or large the percentage. I created scorecards so that users would see the numbers first and could skim easily.
Results from first iteration
Splitting Bulk Action dashboard into two tabs: Details and Progress
Throughout the layout ideation and iterations, I received feedback from design, engineering, and product throughout and maintained consistent communication. Some of the key pieces of feedback included:
  • Insights from the page are not actionable
  • Progress doesn’t feel clickable or interactive
  • Landing page is not beginner friendly
  • Wording between the bulk action list and detail view is confusing
Feedback
Interactions & Call-to-action
Moving to the second iteration, I addressed the feedback received from the first iteration while moving to a higher fidelity.
Feedback: The insights from the page are not actionable
A key piece of feedback I focused on was giving user next steps when seeing errors and problems come up in the bulk action.
I added functionality so that users can create another action and troubleshoot by directly selecting vehicles from the vehicle list.
In addition, I added functionality to allow users to pause and cancel bulk actions to give users more control over the bulk action even after it starts.
Feedback: Users wanted to focus on vehicles with errors
I wanted to connect the progress visualization status cards with the vehicle list. For example, clicking on “Error” would filter the list to only show vehicles with the “Error” status.
I iterated on the scorecard design so that they were clear to read, consistent with the rest of the design system, and felt clickable for users.
  1. Removed the icon graphics so that they wouldn’t overlap and redefine existing icons in the design system.
  2. Added an “All” card so that the users could return to all vehicles after visiting a specific status.
  3. Used a radio button at the top left so that it felt familiar for users to select an option, similar to a form.
Handoff
Delivering the product
I created the Bulk Action Detail View for 6 types of bulk actions: SIM status update, Change Cellular Configuration, Change Trial Plan, Change Lifecycle Status, Change MSISDN, and Change Rate Plan.
Each of these bulk actions had different timelines, groupings, and rollout processes. I worked closely with product and engineering to flesh out each of the bulk action types.
Takeaways
Next steps
Handing the screens off to development, my work for the Bulk Action Detail View has finally come to an end. If I had more time, I would have done more testing and rounds of feedback targeting specific use cases.
Because of the release deadlines, the version I helped handoff into development is not the final version of these screens. I took out filters, search, and vehicle selection so that the screens could be developed more feasibly within the time frame. However, these features along with stakeholders and alerts will be added in the next release of the management platform. I would also reconsider some design decisions:
  • Are scorecards really the best way to give users a sense of the vehicle status distribution?
  • How does this page relate to the stakeholder approvals flow?
  • What does this page look like before it is approved and fully started?
  • More usability testing: at which points do users need more context or have trouble understanding?
  • More user feedback: how does this compare to interaction patterns users are used to?
Learnings
I’m amazed at how much I learned from these few weeks on this task.
One aspect that isn’t shown directly in my process and results is my improvement in using Figma. Prior to this internship, I had given basic workshops on Figma and using it for UI/UX projects. However, I did not fully utilize the more advanced features such as autolayout, constraints, components. As I did working sessions with the product designers, I started picking up more tricks that made my work expeditiously more efficient. I learned about rule of 8s, using option key when working towards high-fidelity, and how to utilize design systems just by watching and observing the product designers. I believe it is essential to keep learning and picking up new tools or methods as a designer.