An end-to-end design system for urban mobility at scale

Role: Lead UX Designer
Timeline: Nov 2018 — Mar 2019
Engagement: Full-time employment

Situation

A rider boarding a Shuttl bus in a congested corridor needed proof of check-in without reliable Wi-Fi or cellular, while drivers needed a fast confirmation that did not create privacy or noise failures. Ops needed accurate ridership. QR, GPS geofence, Wi-Fi, and cellular approaches failed under real corridor conditions.

My mandate

I owned: Product design for rider, driver, and corporate surfaces; design-system foundation; data-over-sound check-in experience and failure states.
Others owned: Product owned roadmap. Engineering owned audio/data protocol feasibility. Operations owned field rollout.
Final decisions: Product on shipping check-in method; engineering on technical viability.
Team: Lead UX Designer with product, engineering, and ops across B2C and B2B lines.
Authority: Lead UX ownership of system patterns and the check-in journey.
Delivery constraints: Noise, privacy, device diversity, and intermittent connectivity in Delhi-NCR-style corridors.

Critical decision

Situation

Choose a boarding check-in method that works when QR scanning, GPS geofencing, Wi-Fi, and cellular handshakes fail in crowded stops.

Options considered

QR scan at boarding
Failed under queue pressure, damaged codes, and riders without a ready camera flow — slowed boarding.
GPS / geofence auto check-in
Unreliable in dense corridors and multi-bus stops; false positives and missed boardings.
Wi-Fi or cellular handshake with the bus
Congested networks and carrier dead zones made connectivity the failure mode, not the rider.
Data-over-sound check-in — privacy-preserving, network-agnostic

Observation

Field observation of boarding showed network-dependent methods collapsing at peak. Drivers needed a confirmation that did not depend on the rider’s data plan.

Insight

Audio data transfer introduces new failure states (noise, volume, device mic) — but removes dependence on corridor connectivity that the product could not control.

Response

Ship automated data-over-sound passenger check-in with explicit failure recovery, alongside a design system holding rider, driver, and corporate surfaces.

Result

Check-in and live tracking became core differentiators. Company-context scale (1M+ daily riders, 80+ corporates) describes Shuttl’s platform, not a personal KPI.

How the system changed

01
Documented why QR, GPS, Wi-Fi, and cellular failed for riders and drivers in congested stops.
02
Designed data-over-sound check-in with failure states and recovery instead of a happy-path-only demo.
03
Built live tracking across B2C and B2B so riders and fleet managers shared operational truth.
04
Scaled the design system through payroll integration, tiered pricing, and fleet optimisation without fracturing language.

Outcome

operational

Operational: check-in that works offline-ish

Data-over-sound boarding confirmation without depending on Wi-Fi or cellular at the stop.

business

Company context: platform scale

Design system held rider, driver, and corporate surfaces on a platform that served 1M+ daily riders and 80+ enterprise clients.

organisational

Organisational: durable practice

Years of staff-level work established patterns that could scale across consumer and enterprise lines.

Scale figures describe the operating context. Personal contributions are stated separately.

Frames

Rider surface: booking and boarding in one system language — check-in designed for corridor failure, not lab connectivity.
Data-over-sound check-in: network-agnostic confirmation after QR/GPS/Wi-Fi/cellular approaches failed in the field.
Corporate and driver ops: live tracking and fleet views share the same system so B2B and B2C do not diverge.

What I would change now

Lead the case with the check-in failure matrix and field notes as the spine, and keep the design-system overview as supporting context rather than the headline.

Have a similar challenge?

Work with me.

Start a conversation