What shipped

Changelog.

Month by month, in plain terms: what a plant manager can do now that they couldn't before.

September 2026

The app tells a plant manager where to look first, prices every stop in dollars, and adds two-factor sign-in.

  • A daily management board gives the morning meeting a single screen: yesterday against target, the top losses in dollars, and open actions with an owner.
  • The downtime page prices every reason and every stop in dollars, based on the rate and the minutes once each stop is resolved.
  • Home now opens with a “look here now” strip tailored to whoever is viewing it, maintenance, quality, or supervisor, plus a running total of confirmed savings for an owner or general manager.
  • A manager digest is assigned automatically when someone is invited, so a plant manager who never opens the app still gets the numbers by email.
  • A new machine gets a starting baseline on day one, drawn from similar machines already on the plant, instead of waiting weeks for its own history.
  • New accounts start with a simple six-item menu, with more screens appearing as a role needs them, instead of the full app on day one.
  • Signals from a newly connected machine are classified automatically by the shape of their data, counter, state, alarm, load, so most tags need no manual setup.
  • The same alert repeating for one ongoing condition is limited, so one crossed threshold doesn’t flood the alert list with duplicates.
  • System-health alerts, like a gateway that has gone quiet, are now kept separate from production alerts on Home and the alert bell.
  • If a gateway goes quiet, it raises one alert for you and automatically notifies Spall’s support team.
  • Any account can turn on two-factor authentication with an authenticator app and one-time recovery codes. An admin can ask every other admin on the account to enrol.
  • A one-click export bundles every dataset in the account as CSV files, so your data can leave anytime.
  • The wallboard can show a live downtime-cost tile: what today’s stops have cost so far.
  • The language toggle between English and Spanish is now reachable from the app itself.
  • The public API adds rate limits, safe-retry keys (idempotency keys) so a retried request doesn’t double up, and self-service webhook management for anyone integrating from an ERP.
  • A security vulnerability disclosure policy is published on the site.
  • A new page explains what the AI learns about a plant, what it does with what it learns, and what it never does.

August 2026

Ask your factory a question in plain language, carry a physical andon light, and use the app comfortably from a phone.

  • Ask Your Factory answers plain-language questions about your plant, including cost questions, using maintenance history, logbook notes, jobs, quality, and machine health together, not one at a time.
  • A physical andon light can plug into the gateway: a tower light or beacon visible from across the floor that shows machine status, including a “no data” signal instead of just going dark.
  • Push notifications land on an installed phone or tablet, so an andon call or an alert can reach a lock screen without the app open.
  • What used to show as a work order is now called a Job everywhere in the app, matching what the plant floor already calls it.
  • A per-cycle run review shows every part’s cycle time against target across a job and a shift average.
  • The app works on a phone and a tablet: the facility map, cell monitor, and run review each got a layout built for a small screen instead of a shrunk-down desktop view.
  • A searchable Help Center is live, with step-by-step articles for setup and daily use.
  • A developer portal publishes the API reference with a live try-it console, for anyone integrating from an ERP or a warehouse.
  • The downtime threshold, how short a stop has to be before it doesn’t count as downtime, is configurable per plant instead of fixed.
  • One gateway can now read many sensor points at once through an off-the-shelf expansion module, instead of needing a separate device for each point.
  • A Settings hub gathers plant configuration that used to be scattered across separate screens into one place.
  • Every user can be assigned more than one role, so a person who wears two hats on the floor sees the screens for both.
  • Terms of Service are published on the site.

July 2026

The floor comes online, machines connect natively, and the numbers start computing themselves.

  • Live OEE (availability, performance, quality) updates automatically from your machine data, with trend charts instead of a spreadsheet built by hand.
  • Downtime is detected automatically and timestamped, with a two-tap reason prompt for the operator and a chart ranking where the time went, biggest losses first.
  • Production counts and shift summaries update in real time as parts run, no manual tally sheet.
  • Machines connect directly, over the protocol they already speak, OPC-UA, Siemens S7, Allen-Bradley EtherNet/IP, MTConnect, Modbus, REST, and MQTT, or a CSV file if nothing else fits, so most controllers connect without custom integration work.
  • Tag Discovery surveys a newly connected machine and ranks which of its signals are worth watching, instead of a technician guessing at a tag list by hand.
  • Every loss gets a dollar figure, calculated from your own labor, scrap, and energy rates, so the dashboard shows money lost instead of a percentage.
  • Alerts reach you by email when a machine stops, a quality threshold is crossed, or a shift target is missed. Each one tracks who acknowledged it and, if no one did, how it got escalated.
  • Reports export to CSV and Excel on demand, or arrive on a schedule as a shift, daily, or weekly email digest.
  • Defect tracking and first-pass yield turn quality into its own screen, with a chart ranking scrap by cause instead of a paper inspection log.
  • Maintenance work is tracked from acknowledge to assign to repair to close, with average repair time and average time between failures (MTTR and MTBF) computed per machine.
  • A tablet-first operator scoreboard shows target versus actual and pace for the shift, with a locked operator view and a separate manager view.
  • Andon calls route to whoever should answer, with response time tracked from the call to the fix.
  • Changeover time is tracked by product pair, so a changeover that grew doesn’t stay invisible.
  • Energy use per unit and idle power draw show up per machine, using the same connection already pulling production data.
  • A shop-floor kiosk mode puts a live wallboard on any screen, no login required, for a TV bolted to the wall.
  • A facility map shows the whole plant’s machines in one live layout instead of a list.
  • Your data can leave the app on your own terms: an API for developers, a live feed into Excel, and automatic pushes to other systems (webhooks).
  • Setup runs through a guided wizard: connect a gateway, pick a machine, test the connection, and see live data the same day.
  • Operator-facing screens are available in Spanish as well as English.

See it running on your own machines.