BDRLABVANCOUVER · 49.28°N
BDR-26-02Rev △3CLIENT◍ DELIVERED

KEY-ONE_

A fan-shaped 34-key layout and an integrated trackball, designed around one specific right hand after an accident, so that typing and pointing never require moving off the device.

Keys
34
Key travel
2.80 mm
dial indicator, n = 34
Sustained speed
41 wpm
after three weeks of daily use, ten 2-minute trials
Build
17 d
Brief → Delivered
Revision history
△3 · 12-02
Thumb cluster reprinted at 4° more splay after two weeks of use.
△2 · 11-16
Trackball moved 14 mm inboard; thumb no longer leaves home.
△1 · 11-11
Outer column travel 3.4 → 2.8 mm.
FIG. 1 — BDR-26-02-AREV △3

BDR-26-02-A — reserved. Fan layout set from a scan of the hand it was built for. 34 keys, trackball under the thumb.

Spec — as measured
Mass
480 g · scale, n=1
Reach span
168 mm · thumb to little finger, measured on the user
Parts
96
Disciplines: MEC FAB ELE FRM.
§01 · Brief

What we were asked for.

A viewer of ours lost their right hand below the wrist in an accident and had been using a split keyboard with one half removed, which works but requires constantly moving between keys and a mouse.

The brief was to make a single device where text entry and pointing share the same hand position. There was no budget and no schedule; we did it because it was the most interesting fitting problem we had been offered that year.

§02 · Constraints

The targets, and what we actually hit.

Constraints set at kickoff for BDR-26-02, with the measured result and the method.
ConstraintTargetAchievedHow measured
Every key reachable without wrist translationAll 34All 34observed, three sessions, marked on video
Key travel2.5–3.0 mm2.8 mmdial indicator on every switch
Sustained typing speed after 3 weeks> 35 wpm41 wpmten 2-minute trials, standard text
Pointer and text entry without moving the handRequiredMettrackball under the thumb, chorded modifiers

Targets were set at kickoff, before the first part was cut. A target we missed stays on the sheet with the number we actually got.

§03 · Mechanism

How it moves.

The key wells are laid out on a fan whose radius comes from a scan of the actual hand, not from a standard. Three revisions: the first had too much travel at the outer columns, the second put the trackball where the thumb had to leave the home position, the third is the one in use.

Everything is printed except the plate, which is water-jet 1.5 mm aluminium. Printing the shell meant a revision cycle of about eleven hours, which is the only reason three revisions fit inside seventeen days.

§04 · Electronics

What it runs on.

One 2-layer board, an RP2040, and a PMW3360 sensor for the trackball. Nothing exotic — the entire difficulty of this project was mechanical fit, and the electronics were deliberately chosen to be boring so they never became the problem.

Electronics subsystems on BDR-26-02.
SubsystemSpecification
Board2-layer, 68 × 74 mm, 1 spin
MCURP2040
PointerPMW3360, 12000 CPI
SwitchesLow-profile tactile, 34
FirmwareQMK fork, chorded modifiers
§06 · Result

What it did.

Delivered after seventeen days and three key-well revisions. Sustained 41 wpm after three weeks of daily use, measured over ten two-minute trials.

It has been in daily use since. The two revisions since delivery were both about the thumb cluster, which is the hardest part of any one-handed layout and the part we would expect to keep revising.

17 d2 d4 d4 d5 d2 dBrief2025-11-03Hand scan and first key-well mock2025-11-05Board spin 12025-11-09First typed sentence2025-11-13Key-well revision 32025-11-18Delivered2025-11-20

Elapsed days between recorded milestones, drawn to scale from the dates. The overall dimension runs from the brief to the last entry.

§07 · Revisions

What changed since.

Revision history for sheet BDR-26-02.
RevDateChangeBy
△32025-12-02Thumb cluster reprinted at 4° more splay after two weeks of use.DZ
△22025-11-16Trackball moved 14 mm inboard; thumb no longer leaves home.DZ
△12025-11-11Outer column travel 3.4 → 2.8 mm.DZ

What we would do next

The chorded modifiers are learnable but they are the thing that took longest to internalise. A small display showing the active layer would probably have shortened that.

We would scan the hand in two positions rather than one. The fan radius is right for a relaxed hand and slightly wrong when reaching.

§08 · Record

Who built it, and when.

Record — as filed
Entry
CLIENT
Client
An individual, self-funded
Status
◍ DELIVERED
Brief
2025-11-03
Demo
2025-11-20
Elapsed
17 d
Built by
DZ
Disciplines on this sheet
MEC
Mechanical design
FAB
Fabrication and finishing
ELE
Electronics and PCB
FRM
Firmware
MEC · FAB · ELE · FRM
§09 · Commission

Send us the thing that is not working.

We built KEY-ONE in 17 days, brief to demo. Send us the problem you have and we will tell you what it takes.
Send us the problem →
Sheet BDR-26-02 · Rev △3 · 2025-12-02