Product

A local system that turns your facility's data into meaning.

Three layers in a single system — real-time monitoring, facility memory, and a growing data asset. None of them depend on the internet.

Real-time monitoring

Listens to sensors second by second. Raises alarms on threshold breaches; with machine learning, it catches behavior that deviates from each asset's own normal.

Facility memory

Stores high-resolution data in your facility for years. 1 year raw, 3 years minute-based, unlimited hourly — go back in time whenever you need.

Your data asset

The data you accumulate becomes a solid foundation for your sustainability reporting, energy-efficiency studies, and R&D projects.

WATCHES, NEVER INTERVENES It never touches your process, automation, or machinery in any way.
How It Works

Smart monitoring in three steps.

This entire process happens in a single local system, without leaving your facility.

01 · CONNECTS

Reads the data

Reads data from your sensors and automation system. Uses industry-standard communication; connects in parallel to your existing infrastructure.

02 · INTERPRETS

Produces meaning

Turns raw data into meaningful information with equipment templates, KPI calculations, and machine-learning models. Learns each asset's own normal.

03 · DRIVES ACTION

Guides the operator

Alarms, the maintenance checklist, and the technical assistant are presented together. When you want to look back, all the data is already there.

Anomaly Detection

Catching a failure before it happens.

  • Beyond the classic threshold alarm. Machine-learning-based detection learns each asset's own normal: it says, "this pump shouldn't be running like this at this hour."
  • A meaningful alarm lifecycle. Compliant with the ISA-18.2 standard: active, acknowledged, cleared. It can be silenced, tracked, and reported.
  • Context, not noise. Every alarm comes with context: which asset it came from, how long it has persisted, and its urgency level.
  • custos.local / kpi
    Trend chart with an anomaly caught
    Operator Experience

    Smart alarms, solutions, and an advisor in one place.

    Knowledge lives in the system, not in individuals. Even when the team changes, the knowledge isn't lost.

    Alarm

    Not just "something happened" — which equipment, for how long, how serious, and whether there's a precedent.

    CRITICAL Feed pump exceeded its current limit
    WARNING Chiller condenser outlet pressure rising
    INFO Zone-3 circulation pump engaged

    Maintenance checklist

    After an alarm, the steps for the technician are presented in order: "First check X, then measure Y…"

    • Filter status checked
    • Valve position verified
    • Pressure measurement

    Technical assistant

    Natural-language Q&A; the system answers from its own knowledge base.

    "Why did AHU-2 raise a warning?"
    AHU-2 needs a filter check — differential pressure exceeded the threshold.
    "What is the filter replacement interval?"
    Data Asset

    Today's monitoring tool, tomorrow's data source.

    Raw Data
    365days

    Raw data at second-level resolution. Keeps every single reading.

    Minute
    3years

    Minute-level summaries. The ideal resolution for trend analysis.

    Hourly

    Hourly summaries kept without limit. For looking back.

    Energy efficiency

    Comprehensive analysis of your consumption patterns, with optimization opportunities.

    ESG / CBAM reports

    An evidence-based data foundation for sustainability reporting.

    Equipment lifecycle analysis

    Which equipment has worn how much, and when it should be replaced.

    Failure prediction models

    Predicting future failures from past failure signals.

    Features

    What's inside Custos?

    01

    Automatic sensor discovery

    Connect the device and the system recognizes active sensors on its own.

    02

    Equipment template library

    HVAC, pump, motor, compressor, cooling, AHU, chiller, FCU.

    03

    KPI calculation engine

    Template-based performance indicators are generated automatically.

    04

    ML-based anomaly detection

    Each asset learns its own normal; deviations are caught.

    05

    ISA-18.2 alarm lifecycle

    Active, acknowledged, cleared — full traceability.

    06

    Maintenance module & checklist

    Periodic maintenance calendar, step-by-step instructions after an alarm.

    07

    Technical assistant chatbot

    Natural-language Q&A; the system answers from its own data.

    08

    Long-term trend with unlimited zoom

    From seconds to years, navigate the past from a single interface.

    09

    Automatic monthly archive (open format)

    Vendor-independent, readable for years to come.

    Contact

    Let's see the value it brings to your operation.

    Let's assess how Custos fits your equipment and your needs in a preliminary call.