Allocate tasks by input capability and screen size

What it is

Searching for a film with a TV remote means walking a focus cursor across an on-screen keyboard; summoning the phone as a keyboard makes the same search finishable. Allocating by input capability and screen size splits one job onto the device that can actually do that step: the large screen for watching, the small one with a keyboard or camera for entry and capture. A video edit with the timeline on a computer, a pickup shot from the phone camera, and a frame retouched with a stylus on a tablet is the same rule. This is not borrowing one device’s hardware as a peripheral of another. It is parking steps on different devices. It is also not the fixed watch-and-phone pair; any complementary set counts.

Why it happens

Input bandwidth and display bandwidth do not line up. A television ten feet away has almost no usable text entry and the largest shared view; a phone has a fast keyboard and a camera on the body, and no room for a timeline. People already work this complementarity: shoot on the phone, retouch on a slateful tablet, export on a computer. Binding the whole task to one form factor forces high-cost steps onto the wrong device—typing on the TV, aligning tracks on a phone, holding a laptop as a viewfinder. The unit of allocation is the step, not the app: the same product can expose only the slice that device is good at. Screen size sets readable density; the input apparatus sets tolerable precision. Scaling a layout by resolution does not fix a step that should never happen on that device.

Studying it

Multi-device ecology work uses diaries, household observation, and device logs to ask which step happened on which device, not how satisfied people were. In the lab, a complementary task that needs both large-screen viewing and precise input (find a title on the TV, type it on a phone; align a desktop timeline, shoot an insert on another device), comparing “all on one device” with a split by capability.

Independent variables: mapping of steps to devices, distance and whether both are in the room, input type (keyboard, touch, remote, camera, stylus). Dependent variables: completion time, errors, abandonment, spontaneous moves of a step onto another device.

Ecology already shows people splitting steps; a lab that only hands over one device will not measure division. Do not collapse “phone types for the TV” with “tablet stylus draws into the computer desktop”—the latter is input borrowing. A division study should keep each device operating its own interface.

Where it stops holding

With only one device present, there is no partner; the task must degrade on that device rather than assuming a keyboard nearby. Location-bound settings (a car, an operating room, a counter) cannot be carried into another role. Assistive technology may live on one device; moving the input step to another can dismantle a working setup. Highly confidential steps should not migrate to a shared screen just because it is larger.

Applying it

• List each step’s demand on pointing precision and screen area. Put dense watching on the large display; put text, precise selection, and capture on the device that has a keyboard, pointer, or camera.
• On a device that is bad at a step, do not ship a shrunken version: TV search should wake a nearby phone keyboard, not a slightly larger on-screen grid.
• Ends of the same account may expose different slices, matching that device’s capability, rather than stuffing the desktop feature tree into the phone.
• Verify by completing a title search on a TV with only a remote, then allowing nearby phone input, and comparing time and abandonment. If the phone path is not clearly faster, the step was not split by input capability.

Related

• Within the group: K8.03.2 The division of labor must be predictable · K8.03.3 Missing capabilities need explanation, not silent hiding
• Adjacent: K4.07 Division of labor and handoff with the phone · K5.06 Difficulty of text entry · K8.08 Borrowing input across devices
• Search terms: complementary devices · multi-device ecology · task allocation