Dashboard built for 40km/h.
India’s delivery riders run 30+ orders a day with a phone taped to the handlebar. I led the UX and interaction design for the dashboard that replaced it.
- Role
- Design Consultant (UX/UI)
- Team
- 1 Product Lead, 2 Senior UX, 3 stakeholders, me
- Duration
- 6 months · Jun 2023 – Jul 2024
- Company
- Sharp × Zuge Electric
- Platform
- 7-inch TFT touchscreen
- Live
- zugeelectric.com
At 40km/h, a two-second glance is 22 metres travelled blind.
Three apps open at once: navigation on one, orders on another, earnings on a third. Every notification was a reason to look down. Every glance was a moment of blindness at speed.
the problem
Every rider was doing this job with a phone taped to the handlebar.
Across 20+ ride-alongs, the same patterns kept surfacing. None of them were about taste. All of them were about safety.
Checked their phone 5+ times per delivery.
Had a near-miss from a glance down at speed.
Felt range anxiety in the middle of a shift.
Wanted a built-in display over a taped-on phone.
research
How might we give riders back the space to just ride?
20+ interviews across Swiggy, Zomato, and Zepto riders. Heuristic teardowns of the Ola S1 Pro, Ather 450X, and TVS iQube. Lo-fi and hi-fi prototypes tested in real traffic, not at a desk. Four riders became recurring design partners.
form studies: finding the screen’s silhouette
what didn’t work
The obvious layout was the wrong one.
Split-screen layout
✕ didn't workMap and order, side by side. The logic was airtight: riders needed both, so show both. But on a 7-inch screen at 40km/h, neither half was glanceable. The map was too small, the text too cramped, and riders squinted, the exact behaviour we were trying to kill.
Layered overlays
✓ shippedNavigation owns the full screen. Order, battery, and alerts surface as minimal overlays, present when needed, gone when not. The overlay version scored far higher on glance time and comprehension. That killed split-screen entirely.
the dashboard
Navigation owns the screen. Everything else waits its turn.
“No tap, no app switch?”
He’d accepted app-switching as part of the cost of the work. So navigation now auto-loads the moment an order is accepted, zero extra taps. Order, battery, and calls arrive as minimal overlays: present when needed, invisible when not. Removal was the work, the earnings tracker, in-app messaging, and text-heavy panels were all cut for safety.
the overlays, as components
Battery & range
Incoming order
Rider location
the dashboard, across states
Lock screen
On the road: speed, nav & music
Ride mode
Parked
Settings
Accessibility
in context
Final screens, on the bike.
Rendered on the real chassis, across the colourways riders actually buy. The hierarchy that held up at a desk had to hold up here too: at a glance, in sun, mid-shift.
Turn-by-turn at 48 km/h
Settings & vehicle health
Navigation, normal mode
Parked · CO₂ avoided
the outcome
Less looking down, more riding.
0%
less phone use on the road
0%
faster task completion
0M+
gig riders on the platform
45 → 87
rider satisfaction score
3 → 1
apps a rider juggles
7-inch
the entire canvas
reflection
What I’d carry into the next one.
Removal is the work
The shipped dashboard had half the elements of the first prototype. The earnings tracker, in-app messaging, text-heavy overlays, all cut. Every removal was a safety decision dressed as a design one.
Test where the work happens
Split-screen looked fine at a desk. It only fell apart at 40km/h in real traffic. Testing in the field, not the studio, is the only reason we caught it.
Constraints clarify
A 7-inch screen at 40km/h isn't a limitation, it's the entire design question. Hierarchy, touch targets, glance time: every call was made inside that one constraint.