2024
Designing a Controlled Notification Architecture for Multi-Source Fleet Ecosystems
Context
Fleet managers were drowning in alerts. Every trigger — from battery failures to routine maintenance logs — landed in the same undifferentiated feed. There was no way to tell a critical alert from a routine one, so managers either missed urgent updates or tuned out the feed entirely. My job was to bring order to a system with no existing hierarchy, no filtering logic, and no owner.
Project: Fleet Management Platform
Role: UX Designer (Interaction Design)
Focus Area: Notification enhancement
Duration: 2.5 weeks
The Challenge
Fleet managers were experiencing "information overload," struggling to distinguish between routine updates and high-priority alerts. With no existing way to filter or manage the influx of data, critical vehicle notifications were frequently overlooked. The challenge was to design a sophisticated notification architecture that balanced high-volume data delivery with user control and visual clarity.
My Role
As an Interaction Designer, describe about your role.

How it stated
Analysis of existing Notification trigger points: Interviewing experienced technical architect, we traced every notification trigger point across the platform, and ran a card sort with the PM and Tech Lead to group them by source and urgency — this gave us a shared model of what the system was actually sending, not just what it looked like on screen.
From there, I made three calls that shaped the rest of the work:
Categorize, don't just prioritize
Move the bell, not the mental model.
Delay the ask, protect the signal.
Categorize, don't just prioritize. Time-sensitive alerts stayed instant; everything else got batched into digest summaries, cutting noise without hiding information.
Move the bell, not the mental model. I relocated notifications from the tab bar to the app bar after benchmarking intranet and OS-level patterns (iOS, Android, Windows, macOS) — matching a convention users already carry with them, instead of teaching a new one.
Delay the ask, protect the signal. Instead of front-loading preference setup during onboarding, I introduced contextual tooltips after a user had seen a few notifications — so they understood what they were customizing before being asked to customize it. Critical alerts like battery failure stayed unskippable regardless of preference.
Empowering Notification Control: Research says 55% of users found notification overwhelm as a reason for primary digital detoxes and 71% users uninstall apps due to excessive alerts.
Keeping that in mind we implemented granular user controls where users can now customize notification types, channels, and frequency (instant, daily/weekly digest), while ensuring critical alerts, like battery charge failures, remain unskippable.
Reusable Component Design: To keep this sustainable past launch, I designed the notification module as a single reusable component rather than a set of one-off variants, and documented floating notification behavior separately for iOS and Android since the two platforms don't share the same rules. I handed this off with full navigation flows and UI specs, written so PMs, BAs, and QA could use the same reference without re-explaining the logic to each team.
Streamlined Developer Handoff: Handoff documentation closed the design-to-dev communication gaps, giving four teams (Dev, BA, PM, QA) a single source of truth instead of four separate conversations.
Action
Image description


Heading or headline
Analysis of existing Notification trigger points: Interviewing experienced technical architect, we traced every notification trigger point across the platform, and ran a card sort with the PM and Tech Lead to group them by source and urgency — this gave us a shared model of what the system was actually sending, not just what it looked like on screen.
Image description
Discovery phase involved understanding the eco-system, applications communications, loop holes, and technology and framework used. Key challenges and limitation of technology was the primary need to deliver the phase wise releases of product.
Heading or headline

Detects the vehicle type on scan, and presents relevant inspection templates as per vehicle type.
Design phase. I tried to showcases the key interactions that drives the journey of scanning the bar/qr code to vehicle type identification and presenting the inspection templates based on vehicle type. Every interactions has a specific reason in placing the ui-element in right place.
Heading or headline
Scan code was available near driver side door.
Clearly communicating in input label helped inspectors finding the scan code sticker.

Short description goes here that talks about the below content.
Heading or headline
Short description goes here that talks about the below content.
Heading or headline
Scan feature helped inspector to quickly scan the code and get the vehicle details. This helped avoiding manul entry of 16 digit vehicle identification number.

-
Fleet managers gained granular control over notification type, channel, and frequency — without losing visibility into what actually matters.
-
Critical alerts (e.g., battery failure) remain guaranteed, unskippable, and visually distinct from routine updates.
-
One reusable component replaced what would have been multiple redundant patterns, reducing future build and maintenance cost.
-
Handoff documentation closed the design-to-dev gap, giving four teams (Dev, BA, PM, QA) a single source of truth instead of four separate conversations.
Next: Exploring GenAI-based summarization to auto-prioritize alerts by real-time context, cutting cognitive load further as fleet size scales.


