Industrial & Machine Data Intelligence

01

INDUSTRIAL GROUND

Start with signal identity, machine identity and governed time.

PLC tags, SCADA events, historian records, sensor readings, counters, alarms, fault codes, setpoints, mode changes and cycle events become useful when their source, semantics and clocks remain explicit.

PLC tags SCADA events Historian records Sensor readings Machine identity Alarm events Fault codes Cycle events Setpoints Sampling intervals

02

SEMANTIC DISCIPLINE

A signal is not a state. An alarm is not a failure.

Raw telemetry contributes Evidence. Machine state requires explicit semantics, transition rules and governed chronology. Alarm timing can establish association without establishing root cause.

SIGNAL

An observed value or event

Tags, measurements, status words and counters remain explicit technical observations.

STATE

A reconstructed technical condition

Running, blocked, starved, faulted or idle are established through governed signal semantics and transition rules.

CAUSE

A separate governed claim

Temporal order, alarm occurrence and state transition do not by themselves prove why a machine entered a state.

03

QUESTIONS

Machine questions answered from Ground.

What state was this machine actually in?

Reconstruct state from explicit signals, events and transition rules.

Which signals established that state?

Trace the reconstructed result back to source tags and historian records.

Where did this cycle spend its time?

Separate active processing, waiting, blocked time and inter-cycle time.

Which alarms preceded the stop?

Reconstruct alarm and fault chronology without promoting sequence into cause.

How long was the machine blocked or starved?

Measure governed state intervals over a defined machine population.

What cannot yet be established?

Missing signals, ambiguous transitions and unsupported predictive conclusions stay visible.

04

MEASURES

Define the machine population, clock and event boundaries behind every measure.

Cycle time

Cycle start → qualifying cycle end

Governed elapsed time between explicit start and end semantics for a defined cycle population.

Ground: machine · cycle identity · start event · end event · clock

State occupancy

Qualifying time in a defined machine state

Governed duration attributed to running, idle, blocked, starved, faulted or another explicitly defined state.

Ground: state transitions · machine identity · governed clock

Downtime

Established unavailable interval

Governed elapsed time during which the defined machine population is established as unavailable.

Ground: machine · fault or stop state · restoration transition

05

SIGNAL TO STATE · ILLUSTRATIVE EXAMPLE — SYNTHETIC DATA

Reconstruct machine state instead of reading one tag as truth.

RAW SIGNAL Motor current rises Historian tag I_MOTOR_07

SEMANTIC EVENT Drive active Rule-qualified event

STATE TRANSITION READY → RUNNING Governed transition

MACHINE STATE RUNNING Traceable reconstructed result

Synthetic example. Signals contribute Evidence; explicit semantics establish the reconstructed state.

06

ALARM & FAULT CHRONOLOGY · ILLUSTRATIVE EXAMPLE — SYNTHETIC DATA

Follow the stop through actual event chronology.

14:03:11 Alarm raised Pressure low

14:03:14 Fault observed Drive fault F-27

14:03:15 State changed RUNNING → FAULTED

14:09:42 Intervention Fault reset

14:10:08 Returned to service READY

Alarm and fault chronology establishes timing and association. Root cause remains a separate claim.

07

CYCLE INTELLIGENCE · ILLUSTRATIVE EXAMPLE — SYNTHETIC DATA

See where the machine cycle actually went.

ACTIVE PROCESSING

41.2 s Defined active machine interval

WAITING

8.4 s Qualifying wait inside the cycle

BLOCKED

6.7 s Downstream condition prevents progression

TOTAL CYCLE

56.3 s Explicit start → explicit completion

A cycle duration is meaningful only when its start, end and internal state semantics are governed.

08

DATA SUFFICIENCY

See what the available machine Ground can establish.

Machine state

Signal semantics and transition rules close the reconstruction.

AVAILABLE

The selected machine has governed signal and transition Evidence sufficient to reconstruct state.

Cycle time

Some cycle boundaries are explicit; others remain ambiguous.

PARTIAL

Cycle measures remain bounded to the population with established start and end semantics.

Root cause

Telemetry establishes chronology, not causal mechanism.

NOT_ESTABLISHED

Alarm order, state transitions and fault events do not by themselves establish why the failure occurred.

09

WHY THIS MACHINE STATE?

Trace reconstructed state back to source telemetry.

RESULT Machine State

TRANSITION State Transition

EVENT Semantic Event

SIGNAL PLC Tag

GROUND Historian Record

State, cycle and downtime conclusions remain connected to the source telemetry, time semantics and deterministic rules that establish them.

10

BOUNDARY

Machine state is not root cause or control authority.

Industrial & Machine Data Intelligence may establish signal chronology, machine state, state transitions, cycle boundaries, alarm chronology, fault chronology, downtime and state occupancy.

It does not silently infer root cause, failure probability, remaining useful life, optimal process settings, automatic maintenance action, operator performance, product quality absent quality Ground or autonomous machine control.

11

DAILY DETERMINISTIC INTELLIGENCE

Search, compare, reconstruct and trace machine state without requiring AI.

SOURCE AUTHORITY

Industrial systems remain authoritative.

PLCs, SCADA systems, historians, gateways and engineering systems retain ownership of the records and signals they establish.

TRACE STATE

Follow machine state to source.

Navigate from a reconstructed state through transitions and semantic events to the exact signal or historian record beneath it.

AI OPTIONAL IN DAILY USE

CUSTOMER-CONTROLLED COMPUTE

Deterministic machine intelligence can operate across local, on-premise, private or distributed telemetry without requiring a generative model or mandatory radius19-hosted SaaS runtime.

PRODUCT BOUNDARY

Observation is not industrial control.

The product does not establish automatic source-data correctness, automatic engineering correctness, guaranteed anomaly causality, safety-control authority, autonomous process control, automatic setpoint decisions or operational writeback. It is not a generative copilot for ordinary daily use.

12

SOFTWARE LICENSE

Deploy Industrial & Machine Data Intelligence in customer-controlled infrastructure.

Industrial & Machine Data Intelligence is offered as software under a direct licensing and delivery relationship. Technical sources, signal semantics, reconstruction rules, machine topology and deployment context determine the bounded product configuration.

Discuss Industrial & Machine Data Intelligence with radius19 ← Back to Products

Reconstruct machine state from signals and time.