Home / What it finds What it finds

The questions Artekis answers from the process data you already keep.

Each analysis below opens with real findings from plants running Artekis. The thresholds and the setup detail are on the pages for your engineers.

What it is

A module inside your Ignition gateway

It watches every measurement against its own normal, reviews your alarms, explains stops, checks the gateway, and answers questions in plain English. Every finding lands on one list in Perspective, in plain words, with the machine it is about, what to do now and the evidence behind it. Each change is a recommendation; your engineers decide what to apply.

Finds what nobody knew to askA slow climb, a stuck reading or a pump losing flow, each judged against that machine’s own normal for what it is doing right now.
Explains the stopThe minutes before a stop on one page: the first alarm, the change someone made, the likely cause and what to check first.
Cleans up the alarm listWhich alarms waste your operators’ time, and the one setting that fixes each, written in Ignition’s own property names.
Your engineers decideThey confirm or dismiss each finding, with a name on every verdict, and a finding closes itself when the data shows the fix held.

At a glance

Six analyses, each with its own question.

Each one runs on its own schedule or trigger, reads a set span of history, and hands you one kind of answer. The pictures use example data. Real findings from plants running Artekis follow below.

MONTUEWEDTHU FRISATSUN Alarm HealthGateway HealthMES Watch Anomaly WatchIncidentAsk weeklydaily, 06:00daily, 06:00 hourlyon an alarmon a question Line 2 downhigh tempLine 2 down ? ? ? ?
Alarm Health

Which alarms are wasting the operators' attention, and how do I fix each one?

Alarm activations, last 7 days, worst first Pump3 Low FlowTank2 Hi LevelMixer Fault Comp1 Hi TempValve8 FailFan2 Off 420281200 604019 The top 3 alarms make up 88% of the load. Each comes with a fix: deadband, on-delay, retire.
Starts
Weekly, or on demand
Reads
Alarm journal, last 7 days
You get
Problem alarms, ISA-18.2 scores, one fix each
Real findings ↓
Gateway Health

What got worse on the gateway since yesterday?

Last 24 h. The gray band is normal. Log errors S&F backlog Failing scripts Connections Devices Bad-quality tags WORSE WORSE okokokok
Starts
Daily
Reads
The gateway itself, last 24 hours
You get
Only what regressed, as a morning digest
Real findings ↓
Incident

What most likely caused this stop?

The 30 minutes before the alarm, on the asset and its neighbors −30−20−100 min Line 2 down 1 2 3 1 Hydraulic pressure fell 18% at −12 min 2 Coolant valve stuck open at −22 min 3 Recipe changed at −18 min Ranked most to least likely. Each one links to its trend.
Starts
An alarm goes active, or a script reports a line down
Reads
30 min before it (adjustable)
You get
Ranked likely causes with evidence
Real findings ↓
Ask

What does the gateway say about my question?

Starts
Someone asks, in the Ask panel or from a script
Reads
Live gateway data for one plant area
You get
An answer where each claim cites its tag or alarm
Real findings ↓
Anomaly Watch

Which measurements are acting unlike themselves?

Oil temp while Running, last 24 h, vs. its own history normal range for this state drifting up → The five kinds it looks for driftout of bandstuckfast swingsflapping quality
Starts
Hourly or daily
Reads
Last 24 h of history, per machine state
You get
Odd measurements, historian health, and an alarm proposal once you confirm one
Real findings ↓
MES Watch

Where is production quietly losing time or quality?

Jam stops per running hour, Cell 3, last 90 days 01.02.0 0.80.91.7 DaySwingNight Night shift jams about twice as often as the other shifts.
Starts
Daily
Reads
MES data (stops, counts, checks), last 90 days
You get
Stops, slowdowns, waste and drift tied to a shift, product or hour
New in 1.5

One minute each

Each analysis, on a running plant.

Short walkthroughs of each analysis on a real Ignition gateway, with captions. The longer stories below show it on the job.

Anomaly Watch1:35A pasteurizer reading that crept up three quarters of a degree and a press whose oil was warming toward its trip, both caught long before any alarm, and the alarm Artekis drafts once a person confirms it.
Alarm Health1:15A week of the alarm journal measured against ISA-18.2 and EEMUA 191, and each problem alarm with the setting that fixes it.
MES Watch0:55Two months of stop and production records read every morning, and three presses that wait on maintenance far more often on the night shift, with the numbers and the cost.
Ask0:55A plain question about downtime, answered from the plant’s own records with every source listed, and a question it declines.
Gateway Health0:43Every morning, what got worse on the gateway in the last 24 hours, and where each finding can go.

Anomaly Watch

“Which measurement is drifting, and since when?”

Every hour Artekis checks each historized measurement against its own normal for what the machine is doing at the time. A confirmed problem becomes an alarm your engineer can import.

The Anomaly Watch review queue: two drift cards for press DCM-304 intensifier response, each with its trend and Real issue, Not an issue and Known buttons
Press 304’s intensifier response drifting as its shot sleeve wore: each card has the trend, the span shaded, the figures and one-click verdicts.

Anomaly Watch for your engineers →


Incident

“Why did it stop?”

When a high-priority alarm fires or a line stops, Artekis puts the minutes before it on one page and ranks the likely causes.

One Artekis finding for press DCM-302: what was found, why it matters, what to do now, how to prevent it, and the evidence table
Why press 302 stopped: jet-cool flow fell from about 3.9 to 2.5 gpm while the die ran hotter, the plant’s own procedure for what to check first, and every piece of evidence behind it.
From a line stop to the likely cause1:15Press DCM-202 stops. About a minute later, the likely cause and what to check first, from the plant’s own procedure, are on that press’s screen, with the evidence behind them. It works the same when an alarm goes off.

Alarm Health

“Which alarms waste your operators’ time?”

It reads the alarm journal, measures it against the ISA-18.2 and EEMUA 191 benchmarks, and writes the setting that fixes each problem alarm in Ignition’s own property names. EEMUA 191 treats more than 10 alarms in 10 minutes as a flood.

What your alarms say about themselves0:56A safety guard bypassed for three weeks, six in ten alarms marked High, a cooling tower alarm that chatters. Alarm Health finds each one and writes the fix.

Alarm Health for your engineers →


Ask your plant

“Which press lost the most time this week, and why?”

Your supervisors type a question beside the Perspective screens they already use. The answer comes back in minutes and dollars, with the records behind every claim. An answer that cannot cite its source says so instead of guessing.

Ask your plant1:14Scrap jumped on a press with no alarm. The supervisor asks the press, and the answer names the setpoint change, when it was made and the data behind each claim, then looks ahead to the next press heading for trouble.
Which press are the leak-test rejects in machining coming from?
The answerCastings from press DCM-304 while its shot sleeve wore: 3.6% failed leak test, against 1.0 to 1.7% for every other press.
Only possible because the press and the machining cell live in one gateway
Which cell waits on castings the most?
The answerCell ML-3: 28 waiting-on-castings stops and 1,960 minutes in 60 days, against 6 stops on ML-1.
From the downtime table, through a query your team approves
Artekis in a Perspective session: the jet-cool finding on press DCM-302 with what was found, why it matters, what to do now and the evidence table, and the Ask panel answering the finding’s own follow-up question
A die-casting plant running Artekis: the jet-cool finding on press 302 and the Ask panel answering “what is the likely cause, what should I do about it, and how do I check the fix worked?” Click to open full size.

Both answers are from a die-casting plant running Artekis. Ask reads your production database only through named queries your team writes, so IT decides exactly what it can see.


Gateway Health

“What broke on the gateway last night?”

Every day it reports what got worse on the gateway in the last 24 hours: bursts of errors in the logs, failing scripts, a Store and Forward backlog, devices that dropped and tags that went bad.

Next step

Tell us about your gateway and what you want to find out.

A member of our team reads every message and replies, usually the same day.

Contact us →