Nobody has to be awake.
A critical fault lands at 01:47 on a Saturday. Auto Solar Service reads it, translates it, picks the nearest free technician, books the visit and texts them. The first person to hear about it is the technician.
What it costs to do this by hand
Every platform speaks its own language
One condition, five vendor strings. Somebody on your team has to know that GRID-501, SMA-410 and Fault11 are the same problem, and has to know it at 2am.
Out of hours, the alert waits for a person
Either it sits until somebody opens a laptop in the morning, or somebody gets woken for a fault that could have waited. Both are decisions nobody made deliberately.
Labor gets billed from memory
Hours reconstructed on Friday are the hours you can defend, not the hours you worked. The gap between those two is margin, and it only ever goes one way.
Polling adapters pull from the monitoring platform on a schedule, and anything else posts in through a webhook: SCADA, an APM platform, your own middleware.
Alerts arrive when they are raised, not when someone checks.
Every vendor string resolves to one of 64 SSA codes across 9 categories, carrying severity, dispatch priority, expected labor hours and the parts that belong on the truck.
A code SSA has not seen lands in UNK and waits on an owner to map it, once.
Zone primary first. If they are unavailable, the best free backup by travel time plus current load, inside your own distance limit. Then the earliest slot that fits.
Critical goes out at any hour and reflows what it clashes with. Everything else waits for business hours, because waking a technician for a soiling alert is how you stop having technicians.
The technician's device crossing the site geofence opens and closes the labor clock. Trip fee, labor rounded up to the hour at your company rate, after-hours doubled, parts at list.
Anything the system cannot price confidently is held as a draft for a human, never billed on a guess.
Trip fee, labor rounding, the after-hours multiplier and which severities may wake somebody are settings, not our opinion. Fourteen of them, per organization.
Pick a fault and follow it down the page
Everything below this line is computed live from the real code table, the real billing rules and real zip geometry, against a sample fleet in Yuma. Choose any row and the rest of the page re-runs for it.
SSA
title
- Severity
- C
- Dispatch priority
- 1
- Flat labor
- 2.0h
Distances come from zip centroids, 33,791 of them, at your own average travel speed. A technician further out than your backup radius is never considered, so the one who is 144 miles away does not get woken.
Nobody approved any of this. There is no queue waiting for a dispatcher to arrive at eight.
The clock nobody has to remember to start
Labor is not typed in by anybody. The technician's device crossing the geofence opens and closes the session, and those two timestamps are what the invoice bills.
The job cannot be closed until materials are confirmed: standard parts pre-filled from the code, extras added by hand, or an explicit "none used". The mobile API for all of this is live. The technician app build is not shipped yet, which is the next piece of work, not a claim on this page.
One trip fee per visit, never multiplied. Labor split by the clock in your own time zone, rounded up to the hour with a one hour minimum, billed at your company rate, after-hours doubled. Parts at list price. Tax on the subtotal.
An invoice with no clock data, an unpriced material, or a session over twelve hours is held as a draft and never sent.
What it reads, and what it costs
What your customer pays is the arithmetic above, with your numbers in it.
What Auto Solar Service costs you is a conversation, and it is set with the first pilot tenant. There is no price table on this page to be wrong.
This is the work order you built by picking a fault. It goes with the request, so the first conversation starts from your own equipment rather than from a demo.