Orbwell.

Kingsville, Ontario · automation & AI consulting · est. on a workbench

Is AI confusing? Overwhelming?

Not sure whether AI — or automation at all — would actually help your business, or where you'd even start? That is exactly the conversation we have, and the conversation is free. Everything we recommend already runs daily on one real property — 38 scheduled agents, a three-product sensor line, dashboards on five screens; the demo is the house.

Start the free conversation →

38
AI agents on daily schedule
3
hardware products in the line
1 day
CAD to live dashboard (TURF)
0
external requests on this page
01

Services & skills

consulting first · every skill below runs in production today

Orbwell consults on computer automation and applied artificial intelligence — and builds, when building is the right answer.

S-01

Computer automation

The work your team does by hand, done by machines that log what they did.

S-02

AI agents & fleets

Scheduled agents doing real work, with a run recorder that proves what ran.

S-03

Reporting

Standups, desk reports, night audits — from your real data, on schedule.

S-04

Dashboarding

Dashboards on the screens you actually look at — TVs, tablets, kiosks, phones.

S-05

Monitoring & alerting

Watchdogs that compare what fired against what actually got written.

S-06

Rapid prototyping

Parametric CAD to firmware to a live dashboard — in a day (case 03 is the receipt).

S-07

Data pipelines

Catalogs and reports that rebuild themselves from approved source data only.

S-08

Home & facility automation

Home Assistant done properly, with power budgets sized for December.

Engagements run small and fixed-scope: audit, first working build inside a week, instrumentation, handoff with the source. Nothing recurring until something is running.

02

The free conversation

no fee · no funnel · no obligation

Our business model, in one line

The advice is free. You only ever pay for a build.

Bring a process that runs badly, a product you want built, or a thing you wish you could see on one screen. We'll point you in the right direction — what we'd build, what we'd buy, and what we'd leave alone. If it turns into an engagement, good; if not, you still leave with the right information.

hello@orbwell.ca

write to Andrew

03

Case files

supporting evidence · imagery from the working CAD or the bench
CASE 01

The Home Assistant estate

Status in productionStack Home Assistant · Python · ClaudeProof 4 silent failures caught in one audit

One property — house, greenhouse, lawn, theatre, print farm — run by 38 scheduled AI agents on top of Home Assistant: standups, triage, watchdogs, desk reports, a night audit. The interesting part is the instrumentation — a run recorder logs what each agent wrote, and a watchdog compares file timestamps against the scheduler, because “the task fired” and “the work happened” are different claims. A single audit pass caught four agents firing on schedule and writing nothing — one quiet for two weeks.

If an agent can fail silently, it isn't finished. That rule is most of the product.

  • By the numbers
  • Agents on schedule 38
  • Standup + CEO reports daily
  • Watchdog truth source file writes, not logs
  • Night audit independent model cross-review
  • Silent failures caught 4 in one pass
  • Screens driven office TV · theatre · tablets · kiosk
  • Anything public human-approved, per item
CASE 02

GrowV1 — a greenhouse sensor line

Status hardware in the fieldStack OpenSCAD · PETG · ESP32 · solarProof ~2.2 Wh/day node budget

A three-product line designed, printed, and flashed in-house for a backyard greenhouse: the Pod (ESP32-CAM climate + camera node), the Solar Power Box that feeds it, and TURF, a micro soil sensor (case 03). The unglamorous work is power — nodes sleep on a one-hour duty cycle and burn about 2.2 Wh a day, sized for December in Ontario, not a demo in July. Every enclosure is parametric OpenSCAD, mesh-verified before slicing, because a printer will happily spend six hours manufacturing a mistake.

  • Products Pod / Power Box / TURF
  • Node power budget ~2.2 Wh/day, 1-h duty cycle
  • Autonomy, no sun ~12 days on a small pack
  • Pipeline CAD → verified STL → firmware → HA
GrowV1 Pod on the bench, wired build with sensors and camera carrier before the lid goes on
Pod on the bench — wired, pre-lid. A real unit.
GrowV1 Pod Rev J.1 assembled, base plus fan-ready lid, labeled CAD render
Pod Rev J.1 — render from the working CAD.
TURF build number one: three printed parts plus bin electronics, annotated render
TURF build #1 — three prints + bin parts.
CASE 03

TURF — rev B to rev G in a day

Status Rev F built · Rev G drawnStack parametric CAD · mesh QC · ESP32Proof six revisions, one bench, one day

TURF is a micro soil-moisture sensor that installs like an irrigation peg — a Ø64 mm cap carrying a 50×50 solar cell, sitting 10 mm proud so the mower passes straight over it. It went from rev B to rev G in a day: parametric CAD, mesh-verified STL, firmware, and live readings in Home Assistant the same evening. The revisions were earned, not cosmetic — rev A trapped its own battery; rev G swaps radio for a wired RS-485 Modbus variant, drawn and deliberately unprinted until the real probe is measured.

An iteration that costs an hour changes what you're willing to fix.

TURF peg family, annotated render from the working CAD
The peg family — annotated render from the working CAD.
TURF mini peg cutaway showing cell, radio and probe path, annotated
Mini peg, cutaway — cell, radio, probe path.
TURF head-house macro cross-section: o-ring seat, sensing chamber, wire path
Head-house macro — o-ring seat, chamber, wire path.
CASE 04

Put A Hat On It — an AI product pipeline with a storefront attached

Status live · putahatonit.caStack Blender · static site · vision QCProof renders derive from the print files

putahatonit.ca sells 3D-printed hats for robot mowers and vacuums — and is the testbed for a full AI product pipeline: design → render → copy → staged social, with a human approval gate on everything that publishes. The honesty rules are mechanical: renders are built in Blender from the actual print CAD in true filament colours, and AI lifestyle images must pass a QC gate that has already bounced an image missing the product. The catalog rebuilds from approved SKUs only; the site is static on Cloudflare Pages, like this one.

putahatonit.ca →

  • Pipeline rules
  • Renders from real CAD, true filament colour
  • AI imagery allowed, labeled, QC-gated
  • QC gate rules → vision model → human
  • Catalog rebuild automated, approved SKUs only
  • Publishing per-post human approval
  • Hosting Cloudflare Pages, static
Also on the bench: Datsun 260Z Pi telemetry (planned) Marketplace scout (live) Maple propagation program (stratifying) Greenhouse winter build — climate battery + thermal reservoir (scoped)
04

Contact

ask for Andrew · replies come from the engineer

One email — no forms, no chatbot. Tell Andrew what runs badly, and you'll get a straight answer on what we'd build first. The first conversation is free.

hello@orbwell.ca