Gloves are a design constraint, not an edge case
Most shop software is designed on a laptop and then handed to somebody wearing welding gloves. It shows immediately: buttons sized for a mouse pointer, forms that need a keyboard, confirmation dialogs that assume you can tap precisely.
We learned this the specific way. On an earlier version of the kiosk, two workers ended their shifts when they meant to start a break — the two buttons sat twelve pixels apart at identical sizes. The fix was not a confirmation dialog asking "are you sure." It was geometry. End Shift now takes about two thirds of the screen, breaks are separated below a labeled divider, and the whole stack flexes to fill the tablet so nothing scrolls.
The rule that came out of it is now a constraint on every shop-floor surface in FabWise: nothing may require tap precision. Buttons on the kiosk and Workstation have a 200×60 pixel floor — well past the accessibility minimum — because a gloved thumb is not a mouse pointer and pretending otherwise produces bad data, not just frustration.
That is the whole reason a welding crew will actually use it. A time system people fight with gets fed garbage within a week.
Long machine runs don't need babysitting
A lot of welding work is not a series of short, tidy activities. A run takes three hours. A fixture gets set once and produces parts all morning. Somebody is welding out a frame until lunch.
Software that assumes constant interaction breaks on that. FabWise doesn't:
- A task runs as long as the work does. There is no maximum duration, no auto-close after an hour, nothing that assumes a task is a short unit.
- Coverage is informational, never a nag. A shift that is one long task is a perfectly good shift. FabWise reports how much of the paid time is attached to a job, and never turns that into an exception a supervisor has to clear. Same for gaps between tasks.
- The tap is not the only way in. Time can be recorded as a duration after the fact rather than requiring somebody to be standing at the tablet the moment the machine stops. On Professional, that includes telling Foreman what you worked on at the end of the day.
The point is that the shop's rhythm sets the pace, not the software's expectations about how often you should be touching a screen.
What you get out of it
Every hour on the floor is attached to a job as it happens, so the questions that were guesses become arithmetic:
- What did that weldment actually cost in labor? Not the estimate — the hours.
- How did it compare to what we quoted? Quoted labor against actual labor, on the job, while it is still open.
- Where does the time really go? Fit-up versus welding versus grinding and rework, if you track labor codes at that level.
- What should the next one be priced at? A shop that has run ten similar jobs at accurate hours quotes the eleventh from a distribution rather than a gut call.
That is the whole loop: capture accurate time, attach it to the job, and quote sharper because you finally know what the last one took.
The rest of the shop, in the same app
The thing shops tell us they value most is that a worker who is not a computer person has exactly one tool to learn. Clocking in, tagging a job, taking a break, requesting a day off, and seeing their own hours are all the same device and the same few taps.
Around that: a schedule that generates shifts so a clock-in can be judged against something, a payroll CSV at the end of the period, and an admin console you can check from your phone without walking back to the office.