/ modules
The full catalog: what problem each module solves, what it does, and what it is worth.
Showing 14 modules
The whole fleet, at a glance.
No single screen showed the whole fleet's live output; checking status meant opening separate control screens or phoning the sites.
Puts every site on a country-scale live map with time travel and a kiosk mode, and stays up on last-known values even if the source drops — it never serves an error.
Removes the site-by-site phone round and the screen-by-screen hunt.
Know every tag you own.
No fleet-wide catalogue of measurements existed, so configuration drift between sister units and bad data stayed invisible, and inventory work was redone by hand.
Turns the asset hierarchy into one searchable catalogue, benchmarks units against each other, and logs and works data-quality issues — a single source of truth.
Ends the hand-redone inventory work and makes drift visible.
Plant data, out the door in seconds.
Only a handful of enabled engineers could reach plant data; everyone else filed a request and waited, so the data was effectively locked away.
Democratizes data — anyone can search a tag and download any time window as CSV, Excel or an image, with every export written to the audit log.
Removes the "file a request and wait" queue and the dependency on a desktop data add-in.
Every tag on one timeline.
Visualizing plant data and asking when a piece of equipment was last in a given state depended on an expensive, per-seat trending tool, with access limited to a few users.
Multi-tag overlay and boolean state/period search replace that tool and remove its licence cost — in-house and open to everyone.
Removes the dependency on a per-seat trending licence and opens the analysis from a few people to the whole team.
Every screen, one grid.
Operational screens lived in a per-seat display tool — hundreds of separate displays deep, and unable to draw a threshold line on a trend or sync a cursor across charts.
A no-code page builder and viewer over the same data, with the interactions that tool cannot offer and the platform's own threshold and alarm layer.
Consolidates hundreds of scattered operator displays into one structure and removes the need for a per-seat display licence.
Hear a fault before it's a failure.
Vibration and temperature data was read case by case by an expert, usually too late, and the alternative was a costly per-machine licensed monitoring system.
Evaluates existing vibration data using equipment-appropriate ISO 20816 criteria, with Isolation Forest anomaly screening, alarm tracking and FFT analysis.
Delivers predictive maintenance with no extra hardware and no per-machine licence.
Signals into decisions.
A healthy unit can run below its reference curve for months, and that silent loss is caught only late, after the fact, by the classic monthly spreadsheet calculation.
Compares live output to the reference power curve, quantifies the loss in energy terms, and makes it recoverable.
Makes silent production loss visible without waiting for month-end.
Catch drift before it becomes downtime.
A major rotating-equipment failure means an unplanned, long and costly outage, and threshold alarms miss the slow drift that precedes the damage.
Learns each unit’s healthy profile and scores live deviation on a 0–100 scale; pauses scoring during start-up and ramp to reduce false-alarm risk.
Catches unplanned downtime while it is still only drift.
Every site, in three dimensions.
Reviewing a site meant either a physical visit or flat 2D drawings; there was no walkable, geo-accurate 3D view of the terrain and structures.
Opens a geo-referenced 3D twin of the site (terrain + buildings) in one click from the map — for remote inspection and layout planning.
Makes review and planning possible without travelling to the site.
The equipment inside the site — and the sensors that feed the models.
For every AI project the map of which sensor sits on which equipment, under which tag, was rebuilt by hand — weeks of error-prone preparation per model.
Standardizes the asset hierarchy and a KKS-validated, ML-ready sensor registry — the data contract that feeds every model.
Does the sensor mapping once, and makes it reusable across every model instead of redoing it each time.
From a teammate's build to a live platform tab.
The good tools the team builds — 3D twins, plant calculators — arrived as one-off demos with no live data and no shared access, running on a single machine.
One searchable gallery of live team-built tools, each card tagged with its own kind — 3D twin, calculated tool, or learning content.
Turns delivered tools from one-off demos into shared, live capability.
Speak up, watch it ship.
Requests and issues were collected ad hoc over email and word of mouth, went untracked, and most were simply lost.
Collects feedback raised from inside every module into one board with voting and status, prioritizing improvements and lifting adoption.
Removes the loss between a request and the queue — no intake board in between.
See the year before you live it.
The modelling and machine-learning roadmap lived in scattered spreadsheets and slides, so nobody could see what was running, what came next, or who owned it.
One timeline over the whole fleet, with tasks anyone can add, move, re-own or drop.
Keeps the yearly roadmap on one screen the whole team reads the same way.
Yesterday, accounted for.
Asking what happened overnight meant going site by site and database by database, reading events one at a time — so the review either took an hour or did not happen.
Brings the previous day's events into one paged, categorised list, filterable by site, severity, template and status, and downloadable as an Excel workbook carrying the same breakdown.
Brings the morning review down from an hour to a few minutes.
SWAPP