Skip to content
ca es en

← Back to list

Planning and driving are two different brains

A dense screen that works at home with coffee is unreadable in the cab with the sun overhead. The temptation was a toggle. The temptation was wrong.

2 min read

Planning and driving are two different brains. Even when it’s the same person. The temptation was to ship one product with a mode toggle. The temptation was wrong.

The context

When I started designing trip mode, the reflex was to reuse the planner view. Same columns, same buttons, an “on trip” flag that changed what was editable. Conceptually clean: one UI, two states.

Operationally, a joke. The planner screen is brutally dense — map, 9-column table, AI side panel, drag-and-drop between days. All of that because you’re looking at it from 50 cm with your hand on a mouse. In the cab, the same screen becomes a carpet of unreadable pixels. The density that made the planner useful kills the driver.

The decision

Two visually incomparable views:

  • Planning view. Table on the left, map on the right, AI and detail panels that appear on demand. You can edit days, drag them, delete them, add POIs. Normal typography, compact spacing.
  • Trip mode. One big card centred on the screen with today’s destination. Three big buttons below: navigate there, mark as visited, skip to next. No table, no side panel, no density. Type scaled +30% by default, high contrast.

Plus a clear state change underneath: the same route moves through three phases — planned, live, done — in one direction only.

Trip state machine: planned → live → done, one-way transitions; each state has its own allowed events.

Why not a toggle. A toggle invites the user to oscillate. Oscillating while driving is the opposite of what you want. And you don’t want it while planning either — trip-mode simplicity isn’t useful there. Making the states non-commutable (forward-only) forces the mental decision the user already makes: I’m planning now / I’m on the road now / I’m recalling now. The UI confirms it, doesn’t propose it.

What it’s not

To be explicit:

  • Not responsive design. It’s not the same screen behaving differently on mobile. They are two different screens, navigated to via different URLs.
  • Not a skin. Beyond the header and the token system, no components are shared. The trip-mode destination card never appears in the planner.
  • Not a setting. You can’t “force trip mode” from the planner. The planned → live transition is fired explicitly by the user once, before the engine starts.

What’s next

The next post is about why POIs surface on their own and how that avoids the blank-page trap. An empty planner is intimidating. A planner that already has suggestions along the corridor of the route — with everything you don’t like removable — is a conversation.