AWI LABS

Engineering that turns an alert into a fix that holds.

Our engineers find the fault using the data, then work back to the cause, reading the machine against the history in your existing systems and the live data collected. Whether that’s a cause and a method your own fitters can follow, or our engineers on site. Your team keeps the plant running while we deliver the remediation projects.

Start with one critical asset See the loop

SPECIFYINSTALLINTEGRATEDIAGNOSEVERIFY

INTERVENTION RECORD

Four stages, written as the job runs

FORMAT

Signal detected

The asset, the measurement and what moved.

Problem diagnosed

What was found on the machine.

Action taken

What changed physically, and who did it.

Result verified

The same measurement points, read again after the work.

THE LOOP CLOSES Stage four is the one most reporting leaves out.

WHERE LABS SITS

Six steps to closed-loop reliability.

Everyone else stops at the alert. The loop only closes when someone connects the sensors, stands at the machine, does the work, and measures again afterwards. That is what AWI Labs is for.

Process diagram of a six-step reliability loop — connect, monitor, detect, diagnose, act, verify — with a side branch showing an alert handed back to the maintenance team instead of continuing to diagnosis. — THE CLOSED LOOPThe same six steps drawn as one loop, with the branch where an alert is handed back to the maintenance team and the loop is left open.

ON SITE

Condition monitoring and fault diagnosis starting with critical assets.

Condition monitoring is measurements taken on a running machine. For rotating machinery, vibration, surface temperature and motor current are a practical starting point. We select additional measurements around the asset, its operating conditions and the fault we need to understand.

Drawing of a motor, gearbox and pump on a common baseplate with guards fitted, numbered one to six at the drive-end and non-drive-end bearing housings, the pump housing and the starter cabinet, plus an edge gateway. — WHERE THE SENSORS GOA generic motor, gearbox and pump train showing six measurement points and what each one is mounted to — bearing housings for vibration and surface temperature, one phase conductor for current.

Vibration analysis

Accelerometer readings at the bearing housings, read against the machine’s own history so unbalance, misalignment, looseness and bearing wear are told apart.

Temperature and thermal trending

Bearing housings, motor bodies and drives measured continuously, so a rising trend and its rate of change are visible long before anything is hot to the touch.

Motor current and load

The motor’s own supply used as a sensor: load, duty and current unbalance tracked without stopping the drive or touching the control system.

Root cause failure analysis

Working back from the failed part through alignment, lubrication, load, installation and process until the cause is named — not just the part that broke.

The measurement follows the problem.

Vibration, surface temperature and motor current are a starting point for rotating machinery. Where the application calls for it, we can select from a wider range of sensors or use suitable existing instrumentation. The choice depends on what needs to be understood, where it must be measured and how the reading will be acquired.

Machine condition and operation

What the machine is doing, and how it is bearing up while it does it. A pump that still delivers but draws more current to do it; a machine whose running hours never quite match the production record.

Vibration · electrical current · contact temperature · high-temperature and thermocouple measurements · open/closed status · dry contact and pulse counting · tilt and angle · distance and position · weight and load

Process and utilities

The conditions the machine is working against. Filter restriction showing as a rising differential pressure, a vessel level that explains a duty change, or compressed air disappearing overnight through a leak nobody has found.

Air, water and gas pressure · differential pressure · liquid flow · tank, vessel and bin level · water and liquid leak detection · airflow · water quality

Working environment

Ambient conditions change how a reading should be read. A bearing running warmer through a heatwave is not necessarily a bearing in trouble, and the record should be able to show which it was.

Ambient temperature and humidity · CO₂ and air quality · particulates · environmental noise level · light and illuminance

Mobile assets and wider site

Where a requirement extends past the machine itself. Secondary to the engineering work, but available when a site needs it alongside the reliability programme.

Asset location · motion · people counting · weather · soil and environmental measurements

Table of engineering questions — filter restriction, pump delivery, vessel level, liquid leaks, whether a machine is running, changed ambient conditions — each paired with a measurement to consider and the context needed to interpret it. — SELECTING THE MEASUREMENTVibration, surface temperature and motor current are the usual starting set on rotating machinery; anything beyond that is specified because a question needs it. Each row pairs the question with a measurement to consider and what has to be read alongside it — illustrative selection routes.

SENSOR RANGE, MOUNTING, SAMPLING, COMMUNICATIONS AND INTEGRATION ARE CONFIRMED FOR EACH APPLICATION. AWI IS HARDWARE-NEUTRAL — A DEVICE BEING AVAILABLE DOES NOT MEAN IT IS ALREADY INTEGRATED, OR SUITABLE FOR EVERY INSTALLATION. ENVIRONMENTAL NOISE IS A SITE CONDITION MEASUREMENT, NOT MACHINERY ACOUSTIC DIAGNOSIS.

Specialist techniques brought in when required.

Some jobs need a technique we do not yet run in-house. Rather than leave you to find someone, we bring in a specialist we trust and brief them on what the data already shows.

Coverage matrix pairing six techniques — vibration, temperature, current and load, ultrasonic, oil analysis and thermography — against questions such as bearing distress, oil contamination, leakage and connection heating, each marked in-house or partner-delivered. — CHOOSING THE TECHNIQUEWhich technique can speak to which question: the three measurements run in-house, alongside the specialist routes brought in. A signal is evidence to interpret, not a diagnosis by itself, and the notes set out what each route depends on — mounting, operating context, electrical scope and sample quality.

Ultrasonic inspection

Listening above the audible range for compressed-air and steam leaks, passing valves and dry bearings, marked up at the machine.

Oil and lubricant analysis

Samples drawn from gearboxes and hydraulics, tested for wear metals, water, viscosity and contamination — the oil, or the machine.

Thermographic survey

A thermal camera across motors, drives, panels and steam lines to find the hot joint, blocked cooling path or loose termination.

Condition-monitoring installation and data integration.

Predictive maintenance only works if the right thing is measured, on the right asset, and joined to what you already know about it.

Condition-monitoring specification and installation

Asset survey, measurement points chosen for the failure modes that matter, sensors and gateways fitted and commissioned — vendor-neutral throughout.

Data and system integration

Live readings joined to the history in your CMMS, SCADA, historians, ERP and operator logs, so an asset has one record.

THE RECORD

Every job is written up the same four ways.

Not a case study written afterwards — a record built as the work happens, so the file for an asset shows what was seen, what was done, and what the data said next.

STAGE 01

Signal detected

The asset, the measurement, the date, and the threshold or pattern that moved. The raw trace is kept as it was when it triggered.

STAGE 02

Problem diagnosed

What was found at the machine, what was ruled out, and the evidence behind the conclusion — readings, images, samples, operator accounts.

STAGE 03

Action taken

The intervention carried out, who did it, when the window was, and exactly what changed physically on the asset.

STAGE 04

Result verified

The same measurement points read again after the work and compared with the baseline — the stage most reporting leaves out.

Schematic trend line crossing a baseline band and a configured limit, with a shaded intervention window and a verification window after it, and a colour key for nominal, warning, high, critical and unknown states. — WHAT VERIFICATION LOOKS LIKEThe shape of a closed record: the same sensor and the same measure, read before the intervention and again in the verification window afterwards.

RECORD FORMAT Intervention record — blank template

TEMPLATE — NOT A RESULT

The fields below are deliberately empty. Each one describes what goes in it when a pilot loop closes and the write-up is approved. Nothing on this page is a customer outcome.

FIELD 01 · SIGNAL DETECTED

Asset or process involved

The specific machine, line or process the record covers — named down to the individual unit and its duty, so a reader knows exactly what was running and how hard. Not a department, not a site.

FIELD 02 · SIGNAL DETECTED

Warning signal

What the data showed first, on which measurement, and how it differed from the reference. Coupled with existing data e.g. past interventions, CMMS data, past work orders, parts history, downtime history, SOPs or manufacturer specs.

FIELD 03 · PROBLEM DIAGNOSED

Root cause

The mechanical, electrical or process cause established on site once the asset was inspected or stripped — and how it was confirmed. The symptom the alert described is not the root cause; if the two disagreed, the record says so.

FIELD 04 · ACTION TAKEN

Physical intervention

The work actually carried out on the machine: what was replaced, realigned, rebalanced, re-torqued, relubricated or re-specified — who did it, when, and what the plant had to stop for.

FIELD 05 · RESULT VERIFIED

Downtime avoided or performance recovered

The operational outcome stated in the customer's own terms and drawn from the customer's own records. Where the outcome cannot be evidenced from those records, this field says that plainly rather than carrying an estimate.

FIELD 06 · RESULT VERIFIED

Evidence that the intervention worked

The post-repair measurement set against the agreed baseline, taken on the signal that characterised the fault, and held across the operating regimes the machine normally runs. This is the field the rest of the record answers to — the standard it has to meet is set out below.

TWO HALVES OF THE PARTNERSHIP

AWI Analytics finds and explains the opportunity. AWI Labs helps engineer the result.

The platform reads the data and says what is worth looking at. Labs goes and looks, manages the work, and measures it afterwards.

Explore AWI Analytics

AWI ANALYTICS IS OPEN TO A LIMITED NUMBER OF PILOT PARTNERS — FREE 90-DAY PILOT, BOOKED THROUGH A 30-MINUTE CONVERSATION

Start with one critical asset.

Tell us which machine causes you the most trouble. A 30-minute conversation is enough to work out what should be measured on it, and whether Labs is the right place to begin.

Start with one critical asset Try the calculator

VENDOR-NEUTRALUK-BASEDNO ENTERPRISE MINIMUMS