benikigai/plus-onePublic

Plus One

Prototype in design. Every value is a design target.

Lessons · Variants

Variants

What change would be better, and what does it cost?

6: candidates

Filled marks are on the Pareto frontier; the ring is the recommendation. Vertical bars are the yield interval.

Objectives

Minimise throughput.cycle_time
Minimise cost.cost_per_good_unit

Constraints

takt 60 s (design target)
headcount at most 9
no new station type beyond the brief

The layout variants come from the N+1 factory engine (nplus1labs/factory 9bc6753), a discrete-event simulation that runs this OPD's routing with a ±30% spread on every target time. They are simulated from assumed values; uncalibrated; no unit built. The audio loopback, the RF check and wireless charging are outside the simulation. The other scores are expert estimates.

Variant RTY Cost per good unit Cycle time Heads Pareto

Baseline

Full scorecard (4)

throughput	cycle time	56 s estexpert-estimateEstimated with expert-estimate. Low confidence.slowest station from target cycle times; nothing built
throughput	takt	60 s estexpert-estimateEstimated with expert-estimate. Medium confidence.design target, about 480 units an 8-hour shift
throughput	bottleneck station	`sta:s8`
throughput	headcount	8

—

56 s

estexpert-estimateEstimated with expert-estimate. Low confidence.slowest station from target cycle times; nothing built

—

Bench cell, for the first 100 builder units Recommended
var:bench-cell staffing

Decision 25's layout for the Q1 2027 bench builds. Three benches in a small U, no conveyor, at most three units in the cell. One technician carries each unit through: bench A does S1 to S4, bench B does S5 and S6, bench C does S7 and S8. 100 units in five build days. Layout: line/layouts/bench-cell.json.

merge sta:s1, sta:s2, sta:s3, sta:s4: bench A
merge sta:s5, sta:s6: bench B
merge sta:s7, sta:s8: bench C

Generated by agent · discrete-event · 2026-11-02

Full scorecard (3)

throughput	takt	1,215 s estanalyticEstimated with analytic. Medium confidence.set by the assumed volume: one 8-hour shift less breaks, 21 working days a month
throughput	uph	16.9 good units/h simdiscrete-eventSimulated with discrete-event, 2,000 runs. Low confidence.Evidence: `src:factory-engine`, `src:layout-study`capacity, simulated from assumed values; uncalibrated; no unit built; the technician is the constraint, not a bench; about 17% busy at this volume
throughput	headcount	1

—

U cell, for about 2,000 to 4,500 a month
var:u-cell staffing

Decision 25's bridge between the bench builds and volume. The eight stations as benches in a U with a 2 s hand pass. Two operators: one runs S1 to S6, the other runs S7 and S8. Layout: line/layouts/u-cell.json.

modify sta:s1, sta:s2, sta:s3, sta:s4, sta:s5, sta:s6: one operator walks S1 to S6
modify sta:s7, sta:s8: a second operator runs S7 and S8

Generated by agent · discrete-event · 2026-11-02

Full scorecard (3)

throughput	takt	255.2 s estanalyticEstimated with analytic. Medium confidence.set by the assumed volume: one 8-hour shift less breaks, 21 working days a month
throughput	uph	31.6 good units/h simdiscrete-eventSimulated with discrete-event, 2,000 runs. Low confidence.Evidence: `src:factory-engine`, `src:layout-study`capacity, simulated from assumed values; uncalibrated; no unit built; the operator on S1 to S6 is the constraint
throughput	headcount	2

—

Balanced MadeCell line, for about 10,000 a month
var:madecell-balanced mixed

Decision 24's MadeCell base (a 40 x 8 ft frame, four bays, stations facing out from a central conveyor, operators on the floor outside), balanced for decision 25's volume case. 16 pallets and an overhead restocking rail. S1 and S2 share a workstation, as do S3 and S4; S5 and S6 and S7 each run twice; S8 becomes a cell of three test fixtures. Seven people. Layout: line/layouts/madecell-balanced.json.

merge sta:s1, sta:s2: workstation W1
merge sta:s3, sta:s4: workstation W2
split sta:s5, sta:s6: W3, built twice
split sta:s7: W4, built twice
split sta:s8: a test cell of three fixtures

Generated by agent · discrete-event · 2026-11-02

Full scorecard (4)

throughput	takt	51.03 s estanalyticEstimated with analytic. Medium confidence.set by the assumed volume: one 8-hour shift less breaks, 21 working days a month
throughput	uph	88.8 good units/h simdiscrete-eventSimulated with discrete-event, 2,000 runs. Low confidence.Evidence: `src:factory-engine`, `src:layout-study`capacity, simulated from assumed values; uncalibrated; no unit built; W1 (S1 and S2) is the bottleneck, about 79% busy
throughput	headcount	7
throughput	bottleneck station	`sta:s1`

—

Split S8: test in one station, mark and pack in another
var:split-s8 station-layout

A ninth bench takes the laser mark and pack off S8. S7 and S8 then both sit at 40 s.

add sta:s8: mark and pack move to a new bench after S8

Generated by agent · expert-estimate · 2026-11-01

Full scorecard (3)

throughput	cycle time	40 s estexpert-estimateEstimated with expert-estimate. Low confidence.
throughput	bottleneck station	`sta:s7`
throughput	headcount	9

—

40 s

estexpert-estimateEstimated with expert-estimate. Low confidence.

—

Plus One+ (coming next)
var:plus-one-plus design-change

Coming next. The same board with more parts fitted: light sensor, NFC tag, debug jumper; a clear shell option. No new station: a board-variant BOM at S2 and extra checks at S8, which push S8 past takt unless it is split.

add part:light-sensor, part:nfc-tag, part:debug-jumper: fit the empty footprints
modify sta:s8: add NFC and light-sensor checks

Generated by manual · expert-estimate · 2026-11-01

Full scorecard (3)

throughput	cycle time	62 s estexpert-estimateEstimated with expert-estimate. Low confidence.S8 gains about 6 s of checks
throughput	bottleneck station	`sta:s8`
throughput	headcount	8

—

62 s

estexpert-estimateEstimated with expert-estimate. Low confidence.S8 gains about 6 s of checks

—

Plus One++ (coming next)
var:plus-one-plus-plus routing

Coming next. Plus One+ boxed with a set of backs. Adds S9 after S8, kitting the backs into the box.

add sta:s8: a new S9 after S8 kits the backs into a bigger box

Generated by manual · expert-estimate · 2026-11-01

Full scorecard (3)

throughput	cycle time	62 s estexpert-estimateEstimated with expert-estimate. Low confidence.S9 kitting runs in parallel; S8 stays the bottleneck
throughput	bottleneck station	`sta:s8`
throughput	headcount	9

—

62 s

estexpert-estimateEstimated with expert-estimate. Low confidence.S9 kitting runs in parallel; S8 stays the bottleneck

—

Design intent

Characteristic	Target	Lower	Upper	Where	Priority · owner
size Approved exterior direction (2026-11-01): the diameter of a common magnetic phone charger, so the puck sits flush on one. Replaces the 57 mm concept. The CAD, the magnet ring, the lugs and plus-one-base@1 in this file are still sized for 57 mm and need rework	55.4 mm InferredEngineering judgement, not measured. Low confidence.Evidence: `src:design-brief`design target	—	55.4 mm InferredEngineering judgement, not measured. Low confidence.Evidence: `src:design-brief`design target	Top shell	must industrial-design
thickness Approved exterior direction (2026-11-01): 12.0 mm at most, including the button dome and the base. Replaces 9.5 mm. A 12 mm puck still needs a thinner cell than the stock one; the CAD in this file is still 9.5 mm	12 mm InferredEngineering judgement, not measured. Low confidence.Evidence: `src:design-brief`design target	—	12 mm InferredEngineering judgement, not measured. Low confidence.Evidence: `src:design-brief`design target	Top shell, Base: the interface	must industrial-design
weight Light enough to hang on a phone or a bag	28 g InferredEngineering judgement, not measured. Low confidence.Evidence: `src:design-brief`design target	—	32 g InferredEngineering judgement, not measured. Low confidence.Evidence: `src:design-brief`design target	Base: the interface	should product
button-travel Short, firm click you can find without looking	0.3 mm InferredEngineering judgement, not measured. Low confidence.Evidence: `src:design-brief`design target	0.2 mm InferredEngineering judgement, not measured. Low confidence.Evidence: `src:design-brief`design target	0.4 mm InferredEngineering judgement, not measured. Low confidence.Evidence: `src:design-brief`design target	Plus button	must product
other Magnetic hold on a phone mount. It stays put when a phone in a car mount hits a bump	8 N InferredEngineering judgement, not measured. Low confidence.Evidence: `src:design-brief`design target	6 N InferredEngineering judgement, not measured. Low confidence.Evidence: `src:design-brief`design target	—	Magnet ring	must product
other Lug pull-out. The twist-lock lugs carry the load for clips, lanyards and bikes; the magnets only align	30 N InferredEngineering judgement, not measured. Low confidence.Evidence: `src:design-brief`design target	30 N InferredEngineering judgement, not measured. Low confidence.Evidence: `src:design-brief`design target	—	Base: the interface	must mechanical
other Pogo pad current limit. A shorted module trips safely	300 mA InferredEngineering judgement, not measured. Low confidence.Evidence: `src:design-brief`design target	—	300 mA InferredEngineering judgement, not measured. Low confidence.Evidence: `src:design-brief`design target	Pogo pad	must electrical