Scoped after a free call

Predictive maintenance with IoT sensors and machine learning

Vibration, temperature and current sensors on motors, pumps and compressors can show trouble days or weeks before a failure, but only if someone turns the readings into planned work. We start with simple trend rules that catch obvious drift, add anomaly models trained on your own readings and failure history once there is enough data to test them, and turn every warning into a work order in your maintenance system with readings attached. Because results depend on your data, we scope it after a free call, usually starting with one machine type.

The problem

Why sensor data rarely turns into planned maintenance

Many plants and facility operators have retrofitted sensors and have a dashboard full of readings. What they lack is the step from a reading to a planned job: someone has to notice the trend, decide it matters and raise a work order.

Predictive maintenance is also oversold. A model needs months of readings and a record of what failed and when, and without that it produces confident noise. The honest path starts with rules, collects labelled history and adds machine learning only when it can be tested against real outcomes.

What we build

What our predictive maintenance service includes

Data readiness check

A review of which machines have sensors, how much history exists and whether failures and repairs are recorded.

Trend rules first

Rules for rising vibration, temperature drift or current changes that catch obvious problems from day one.

Failure history capture

A simple way for technicians to record what they found, so future models have labels to learn from.

Anomaly models

Models trained on your own readings, tested against past failures before they are trusted.

Work orders with evidence

Warnings raised as planned work orders in MaintainX or your maintenance system, with charts and the reason.

Model monitoring

Accuracy checked against technician findings, with re-training when machines or patterns change.

Who does what

Automation does the chasing, people make the calls.

Every step is marked: what runs on its own, what a member of staff reviews and what stays with a person who decides.
  1. 01 Automated

    Readings collected

    Sensor readings flow into the time-series store or IoT platform.

  2. 02 Automated

    Rules and model score

    Trend rules and, when ready, an anomaly model score each machine on a schedule.

  3. 03 Automated

    Work order raised

    Warnings above your thresholds raise a planned work order with evidence.

  4. 04 Ops team review

    Technician inspects

    A technician inspects the machine and records what they found.

  5. 05 Engineer or ops lead decides

    Maintenance decisions

    Whether to stop a machine, replace parts or change schedules is decided by the maintenance lead.

How it behaves

How predictive maintenance behaves day to day

What each person sees and what runs in the background, including what happens when something goes wrong.

What the customer or plant sees

  • Fewer surprise breakdowns
  • Maintenance planned around production
  • Clear reasons behind each inspection

What technicians see

  • Planned work orders with charts and the reason
  • A quick form to record what they found
  • Fewer emergency call-outs over time

What the automation does

  • Runs trend rules and model scores on schedule
  • Raises work orders above your thresholds
  • Tracks accuracy against technician findings

Edge cases and fallbacks

Sensor fails or drifts
Flagged as a sensor problem, not a machine problem.
Machine runs a new product or speed
Readings tagged so the model does not treat a new normal as a fault.
Too little history for a model
Rules stay in place and we say so rather than ship a weak model.
Maintenance system API unavailable
Warnings are queued and sent to email until it recovers.

Architecture

Predictive maintenance from sensor to work order at a glance.

The setup we usually start from. We adapt it to the phone system and practice software you already use.
Machine sensors Vibration, temp Time-series store Readings, history Trend rules Day one Anomaly model When data allows Work order MaintainX or CMMS Technician Inspects, records
  • Input
  • Data
  • Logic and services
  • AI step
  • Output
  • People

How it flows

  1. 1 Sensors on each machine send readings to a time-series store or IoT platform.
  2. 2 Trend rules run from day one, and an anomaly model is added once there is enough labelled history.
  3. 3 Warnings above your thresholds raise a planned work order with charts and the reason.
  4. 4 A technician inspects and records the finding, which becomes training data for the next model version.

Project planner

How the project runs, phase by phase.

Typical durations for a first version. Your proposal lists the exact phases, dates and what you sign off at each step.

Predictive maintenance with sensors & machine learning project plan

Typical time to launch: about 11 weeks, then support

  1. 01

    Data review

    1 to 2 weeks

    Machines, sensors, history, failure records and the maintenance system.

  2. 02

    Rules and capture

    About 2 weeks

    Trend rules, work order connection and the technician findings form.

  3. 03

    Baseline model

    2 to 4 weeks

    Feature pipeline and a first model tested against held-back history, where data allows.

  4. 04

    Pilot

    About 3 weeks

    One machine type in production, with warnings compared to technician findings.

  5. 05

    Rollout

    About 1 week

    More machines, run book and monitoring handover.

  6. 06

    Support

    Ongoing

    Accuracy checks, re-training and new machine types.

What it runs on

Real products only, connected through the access each vendor provides. See all integrations

Python, pandas and scikit-learn TimescaleDB, InfluxDB or your IoT platform AWS IoT Core or Azure IoT Hub, where used MaintainX API or your maintenance system n8n for work orders and alerts Grafana or a custom dashboard

Privacy and consent

Rules we design for, with the sources.

General information to plan with, not legal advice. Your counsel or data protection adviser confirms what applies to your practice.

EU: users can access machine data

Under the EU Data Act, which applies from 12 September 2025, users of connected products can access and share the data those products generate, including with repair providers.

Source: digital-strategy.ec.europa.eu

EU: AI used for safety components

If a model ever acts as a safety component rather than a maintenance planning aid, different AI Act duties may apply; ours plans work for people and does not control machines.

Source: artificialintelligenceact.eu

Operator data under contract

Technician names and findings are personal data, so storage and processing are covered by GDPR Article 28 processor terms.

Source: gdpr-info.eu

How we deliver

Agile delivery, in 1 to 2 week sprints.

We deliver in short Agile sprints using Scrum, so you see working software every one or two weeks instead of waiting for a big reveal at the end.

Every sprint

  1. 1 Sprint planning. At the start of each sprint we agree which backlog items we will finish, each with written acceptance criteria you can check.
  2. 2 Daily stand-up. A short internal check-in on progress and blockers. You get a written update in your Slack or Teams channel, not another meeting.
  3. 3 Sprint review and demo. At the end of each sprint we demo what was built on test data or in a shadow run, and you accept it or ask for changes.
  4. 4 Retrospective. The team looks at what went well and what to change in the next sprint, and we share anything that affects you.

How scope stays under control

Product backlog
Every feature, fix and question in one prioritised list you can see at any time.
Acceptance criteria
Plain-language conditions for "done" on each item, agreed before work starts.
Change requests
New ideas go into the backlog with an estimate. Nothing changes scope or price without your written approval.
User acceptance testing (UAT)
Before launch your team tests the real workflow against the acceptance criteria and signs off.

Tools we use

Jira, ClickUp or Trello for the backlogSlack or Microsoft Teams for updatesGitHub for code and version historyLoom or a live call for demos

Project manager / Scrum master

Runs the sprints, keeps the backlog and timeline honest, and is your single point of contact.

Business analyst

Maps how the work runs today, writes the workflow specification and acceptance criteria with your team.

Automation engineer

Builds the n8n workflows, integrations, AI steps and error handling.

Backend developer

Writes custom code where a no-code step is not enough: APIs, data sync, secure storage.

Frontend / mobile developer

Builds forms, booking pages, dashboards, web apps or Android apps when the project needs them.

QA engineer

Tests every workflow against the acceptance criteria, including edge cases and failure paths.

DevOps engineer

Sets up hosting, access, backups, monitoring and alerts, and the deployment pipeline.

Who does what at each phase

Free call
Project manager, business analyst
Understand the problem, check it is worth automating and outline a rough range.
Discovery
Business analyst, automation engineer
Map the current process, systems and data, and agree success measures.
Design
Business analyst, automation engineer, backend developer
Write the workflow specification, data flows, compliance notes and backlog with acceptance criteria.
Sprints
Automation engineer, developers, QA
Build, test and demo in 1 to 2 week sprints; you accept or adjust after each demo.
UAT and launch
QA, DevOps, project manager
Your team tests and signs off, we switch on, monitor closely and train the team.
Support
Automation engineer, DevOps
Stabilisation window, then optional monthly care with monitoring and changes.

A point of contact

One person who can answer questions and approve decisions, with about an hour a week during the project.

Access

Admin or API access to the systems involved, created as separate accounts with the least access needed.

Sample data

Real but non-sensitive records (or de-identified ones) so we can design and test against reality.

Approved wording

Message templates, policies and consent wording (for SMS, your opt-in and opt-out text) reviewed by you.

Vendor agreements

Signed agreements with the software vendors that will handle personal data, such as a BAA or processor contract.

Time for testing

A few hours from the people who will use the system, for the demos and user acceptance testing.

Working system

Workflows, integrations and any apps running in your accounts, under your keys.

Documentation

Workflow specification, data-flow map, run book and exported workflow files.

Training

A walkthrough for the team and a short recorded video for new starters.

Handover

All access, credentials and code handed over; you can keep us or take it in-house.

Stabilisation window

Fixes for anything that does not meet the agreed acceptance criteria after launch.

Support options after launch

  • No plan

    You run it yourselves after the stabilisation window and contact us when you need a change.

  • Monthly care

    Monitoring, error alerts, small changes and updates each month, with an agreed response time.

  • Priority care

    Faster response times for business-critical workflows and a monthly review call.

How a project moves through our team

Free call PM and analyst Discovery Process and data Design Spec and backlog Sprints Build, test, demo Your review Accept or adjust UAT and launch Sign-off, go live Support Monitor and improve
  • Input
  • Logic and services
  • People
  • Output
  • Data

How it flows

  1. 1 A free call with the project manager and business analyst to understand the problem and whether it is worth building.
  2. 2 Discovery and design produce a written specification, data-flow map and a backlog with acceptance criteria.
  3. 3 The team builds in 1 to 2 week sprints; you review a demo at the end of each and accept or adjust.
  4. 4 Your team runs user acceptance testing, we launch, then monitor and improve under a support plan.

FAQ

Predictive maintenance with sensors & machine learning questions, answered.

Sensors record how a machine behaves, a model learns what normal looks like for that machine and flags drift that has preceded failures, and each warning becomes a planned inspection. It works best with months of readings and recorded failures.

It depends on the machine: vibration for rotating equipment, temperature for bearings and motors, current for electrical load, pressure for pumps and compressors. We work with the sensors you have and say plainly if they cannot support the goal.

Trend rules work from day one. A useful anomaly model usually needs several months of readings and a record of failures and repairs. We check this before proposing a model.

It is custom work on your data and maintenance system, using open tools such as Python and scikit-learn. You own the models, pipelines and code.

No. It raises planned work orders with evidence. Stopping machines and changing schedules are decisions for your maintenance lead.

Warnings that led to a real finding, failures that were not warned about, unplanned downtime and planned versus emergency work orders.

Free 20-minute call

Want this for your iot & connected device?

Book a free 20-minute call. We will look at how the work runs today, tell you honestly whether it is worth building and give you a fixed quote if it is.

Book a free 20-minute call