IoT automation and connected devices

From sensor reading to the right person acting on it, without anyone watching a dashboard

Connected devices produce a steady stream of readings, heartbeats and alerts. The value is in what happens next: a ticket opened, a technician texted, a customer told, a bill sent. Make It Flow designs IoT automation that carries each signal to the right system and the right person, with people deciding anything that touches safety, money or a customer promise.

Understand the work

How an IoT or connected device business actually works

IoT covers very different businesses. A hardware startup ships a sensor or a smart product and supports it in the field. A facility or property manager runs building systems, meters and leak sensors across many sites. A cold-chain or logistics operator tracks temperature and location for loads in transit. A grower watches soil moisture and greenhouse climate. A manufacturer retrofits vibration and temperature sensors to older machines to see trouble coming. What they share is a pipeline: devices send data through a gateway or network, a platform stores it, and people need to act on a small share of it.

Connectivity shapes everything. Battery sensors often use low-power networks such as LoRaWAN, NB-IoT or LTE-M, which trade speed for range and battery life. Mains-powered devices may use Wi-Fi, Ethernet or Thread, and smart-building products increasingly use the Matter standard. Higher-bandwidth devices such as cameras or wearables can use 5G RedCap. Messages commonly travel over MQTT or HTTPS into an IoT platform, a time-series database or a vendor cloud, and many teams also keep the data in spreadsheets and email alerts.

The rules are changing too. In the EU, the Cyber Resilience Act brings vulnerability reporting duties from September 2026 and its main security requirements from December 2027, and the Data Act gives users rights to the data their connected products generate. The UK has required no default passwords, a way to report vulnerabilities and a published security update period for consumer connectable products since April 2024. In the US, the FCC's voluntary Cyber Trust Mark label is still being set up. These rules mean more process around devices, not just more data.

Founder or operations lead

Owns the device fleet, the support promise, hardware costs, connectivity contracts and compliance planning.

Firmware and hardware engineers

Build and update device software, investigate faults and decide when a release is ready to roll out.

Platform or data engineer

Runs the gateway, broker, database and dashboards that devices report into, and keeps integrations working.

Facility, site or fleet manager

Responds to alarms, books contractors, reports on energy and uptime, and answers to tenants or clients.

Field technicians

Install, calibrate, repair and replace devices on site, and record what they found and fixed.

Customer support and success

Handles setup questions, offline devices, warranty claims, returns and renewals.

Every day

Threshold alerts and offline devices to triage, support tickets about setup and connectivity, technician visits, shipments to track and new trial or evaluation sign-ups to follow up.

Every week

Maintenance planning from sensor trends, temperature or excursion reports for customers, warranty and RMA decisions, firmware rollout progress and pipeline reviews.

Every month

Usage-based invoices, energy and uptime reports for sites, connectivity and SIM costs checked, security update and vulnerability reviews, and calibration or certification renewals.

Where work gets stuck

The current way of working is full of small detours.

01

Alerts nobody owns

Thresholds fire into a shared inbox or a dashboard that only one person checks. Real problems sit next to noise, and some are only noticed when a customer calls.

02

Too many false alarms

A door left open for a minute, a sensor with a low battery or a brief network drop can all look like an emergency. Teams start ignoring alerts, which is when a real one is missed.

03

Offline devices found too late

A gateway loses power or a SIM runs out of data and the device simply goes quiet. Without a heartbeat check, silence looks the same as everything being fine.

04

Field work disconnected from the data

The alert lives in one system, the work order in another and the technician gets a phone call. Readings, history and photos do not travel with the job.

05

Warranty and returns by email

RMA requests arrive with missing serial numbers, no photos and no fault description. Support staff chase details and check warranty dates by hand.

06

Billing that does not match usage

Usage-based or per-device pricing means exporting data, building a spreadsheet and keying invoices. Mistakes cost money or customer trust.

07

Compliance paperwork arriving all at once

Security update periods, vulnerability handling, user data access and food or product traceability records are becoming obligations, and many teams track them in their heads.

What could flow better

Less chasing. More of the work that matters.

AI agents, assistants, research workflows and integrations can each take on part of the load. Here is where they fit for iot and connected device businesses.

Monitoring and alerts

Turning raw readings into a short list of things a person should act on.

  • Threshold and rate-of-change rules per device type
  • Heartbeat checks that flag offline devices and gateways
  • Grouping repeated alerts into one incident
  • Routing to SMS, Slack, Teams or email by severity and on-call rota

Maintenance and smart buildings

Using trends to plan work before something fails.

  • Predictive maintenance signals from vibration, temperature or run hours
  • Work orders created with readings and history attached
  • Weekly energy and occupancy summaries for each site
  • Calibration and battery replacement reminders

Cold chain and asset tracking

Keeping temperature-sensitive and high-value goods visible.

  • Temperature excursion alerts during storage and transit
  • Excursion logs with start, end, peak and actions taken
  • Arrival and dwell-time notices from location data
  • Records gathered for traceability requests

Warranty, RMA and support

Handling device problems with all the facts in one place.

  • Guided RMA forms with serial number, photos and symptoms
  • Warranty status checked against shipment dates
  • Device logs and last readings attached to the ticket
  • Return labels and replacement shipments queued for approval

Billing and product updates

Charging fairly and keeping customers informed.

  • Usage-based or per-device invoices drafted from telemetry
  • Unusual usage flagged before an invoice goes out
  • Firmware release notes in plain language per customer group
  • Security update and end-of-support notices

Leads and onboarding

Turning interest in a device into a working installation.

  • Trial and evaluation sign-ups qualified and routed
  • Setup checklists sent before the device arrives
  • First data received confirmed, or help offered if it never arrives
  • Follow-up timed to real usage instead of a calendar

Use cases in detail

From trigger to done, step by step.

How it happens today

An email alert lands in a shared inbox. Someone sees it eventually, phones a technician, and the technician arrives without the reading history or the device location.

Systems and information

  • · IoT platform, MQTT broker or vendor cloud
  • · n8n or a similar workflow tool
  • · Help desk or work order system
  • · SMS, Slack or Microsoft Teams
  • · On-call rota in a calendar or sheet

How it could flow

  1. 1 Receive the reading from the IoT platform through MQTT, a webhook or a scheduled API check.
  2. 2 Apply rules the team sets per device type: how far out of range, for how long and whether the reading is rising, so brief blips do not trigger a call-out.
  3. 3 Group repeated alerts from the same device or site into one incident instead of dozens of messages.
  4. 4 Open a ticket in the help desk or work order system with the site, device, recent readings and a link to the chart.
  5. 5 Text or message the on-call technician with a short summary and an accept button, and escalate to the next person if nobody accepts in the agreed time.
  6. 6 Close the loop when the reading returns to normal and the technician records what they did.

Human oversight

The operations lead owns the thresholds and escalation rules and reviews them after false alarms or missed events. Anything involving gas, fire, electrical danger or people at risk follows the site's safety procedures and emergency services, never an automated workflow alone.

What gets easier

The right person hears about a real problem quickly, with the facts needed to fix it, and the team stops drowning in duplicate alerts.

How it happens today

Nobody notices silence. The gap is discovered days later when someone looks at a chart, or when a customer asks why their data stopped.

Systems and information

  • · IoT platform or time-series database
  • · Device inventory with expected reporting intervals
  • · Help desk or CRM
  • · Email and SMS

How it could flow

  1. 1 Check each device's last message time on a schedule, using the expected reporting interval for its type.
  2. 2 Separate a single quiet device from a whole site going dark, which usually points to a gateway, power or network problem.
  3. 3 Look up the battery level and signal strength from the last messages to suggest a likely cause.
  4. 4 Send the customer or site contact a short, friendly check, such as asking whether the router or power was changed.
  5. 5 Open a support ticket or field visit if the device stays offline beyond the agreed window.

Human oversight

Support staff approve the customer message templates and decide when a site visit is worth the cost. The workflow suggests causes but does not claim a diagnosis.

What gets easier

Gaps in data are spotted in hours rather than weeks, and customers see that someone is watching their devices.

How it happens today

Loggers are downloaded after delivery, excursions are found too late to save the load, and putting together a report for a customer or auditor takes hours.

Systems and information

  • · Temperature sensors or data loggers with an API
  • · Telematics or tracking platform
  • · Warehouse or transport management system
  • · Shared drive or document store for reports

How it could flow

  1. 1 Watch live temperature readings and alert the driver, warehouse lead or dispatcher as soon as a reading leaves the range set for that product.
  2. 2 Record each excursion with start time, end time, peak temperature, location and the load or batch it affects.
  3. 3 Ask the responsible person what action was taken, such as moving goods or adjusting the unit, and store their answer with the record.
  4. 4 Produce a per-shipment or per-week report the customer or quality team can review.
  5. 5 Keep the records organised so they can be found quickly if a customer, auditor or regulator asks for them.

Human oversight

Quality staff decide whether goods are still fit for use or must be held, and they own any report sent to a customer or authority. In the US, food traceability rules have their own record requirements and compliance dates, so the workflow supports staff rather than standing in for a compliance programme.

What gets easier

Excursions are caught while there is still time to act, and the paperwork exists the moment someone needs it.

How it happens today

Data is collected but rarely reviewed. Maintenance stays on fixed calendars or happens after a breakdown.

Systems and information

  • · Sensor data in a time-series database or IoT platform
  • · Maintenance management system (CMMS) or work order tool
  • · Spreadsheet or small model service for scoring
  • · Parts inventory

How it could flow

  1. 1 Calculate simple trends from sensor data each day, such as rising vibration, longer run times or higher current for the same work.
  2. 2 Compare each machine with its own history and with similar machines, and where enough history exists, use a small machine learning model to score unusual behaviour.
  3. 3 Draft a maintenance suggestion with the charts, the reason it was flagged and how confident the signal is.
  4. 4 Create a planned work order for the maintenance lead to approve, schedule around production and order parts if needed.
  5. 5 Record what the technician found, so the rules and model can be checked against real outcomes.

Human oversight

Maintenance engineers decide whether to act, and safety-critical equipment follows its normal inspection regime. The model is a prompt to look, not a verdict, and its suggestions are reviewed against what technicians actually find.

What gets easier

Fewer surprise breakdowns, maintenance planned around production, and a record that shows which signals were worth trusting.

How it happens today

The request arrives by email with no serial number. Support asks for photos, checks the order system for the purchase date and copies details into a spreadsheet.

Systems and information

  • · Help desk such as Zendesk, Freshdesk or HubSpot
  • · Order and shipment records
  • · IoT platform device logs
  • · Shipping label service

How it could flow

  1. 1 Offer a guided form asking for serial number, order number, symptoms, photos and what the customer has already tried.
  2. 2 Check the serial number against shipments to confirm warranty status and the product version.
  3. 3 Pull the device's last readings, firmware version and error logs where the platform has them, and attach them to the ticket.
  4. 4 Suggest common fixes from the support knowledge base before a return is approved.
  5. 5 Prepare a return label and replacement order for a support agent to approve.

Human oversight

Support staff approve every return, refund and replacement, and decide on goodwill outside warranty. Customer rights to repair, refund or replacement depend on local consumer law, so staff own any statement about them.

What gets easier

Complete RMA requests on the first contact, faster decisions and a clear record of which faults keep coming back.

How it happens today

At month end someone exports usage, builds a spreadsheet and types invoices into the accounting system.

Systems and information

  • · IoT platform or usage database
  • · Accounting or billing tool such as Xero, QuickBooks or Stripe
  • · Contract and pricing records
  • · Email

How it could flow

  1. 1 Collect usage per customer from the IoT platform or database at the end of each billing period.
  2. 2 Apply the pricing rules agreed in each contract, including minimums, tiers and inactive devices.
  3. 3 Flag anything unusual, such as a sudden spike, a drop to zero or a new device not on the contract, for a person to check.
  4. 4 Create draft invoices in the accounting or billing system with a usage summary attached.
  5. 5 Send approved invoices and answer simple "why is my bill higher?" questions with the customer's own usage chart.

Human oversight

Finance approves invoices before they go out and handles disputes and credits. Pricing rules are written down and changed only by the business.

What gets easier

Invoices that match real usage, sent on time, with fewer disputes and no copy-paste errors.

How it happens today

Release notes are written for engineers, customers find out after their device has changed, and nobody tracks who was told about a security fix.

Systems and information

  • · Code repository or release tool
  • · Device management or over-the-air update platform
  • · CRM or customer list
  • · Email or in-app messaging

How it could flow

  1. 1 Read the release notes or changelog when a release is tagged and draft a plain-language summary for each customer group.
  2. 2 Match the summary to the customers whose devices will receive the update in each rollout stage.
  3. 3 Send the notice before or alongside the rollout, with what changes and whether any action is needed.
  4. 4 Log who was notified and when, which helps show how security updates are communicated.
  5. 5 Collect replies and problem reports after the update and route them to support.

Human oversight

Engineering approves the technical content and the rollout plan. Security wording is reviewed before sending, and the timing of any vulnerability disclosure is decided by people, in line with the rules that apply where the product is sold.

What gets easier

Customers know what changed and why, support is ready for questions, and security communication leaves a record.

How it happens today

Sign-ups land in a form inbox, kits ship without a setup plan, and sales follow up on a fixed date whether the device was ever switched on or not.

Systems and information

  • · Website forms
  • · CRM such as HubSpot or Pipedrive
  • · IoT platform device activation data
  • · Email and calendar booking

How it could flow

  1. 1 Ask a few qualifying questions on sign-up: use case, number of sites or devices, connectivity on site and timeline.
  2. 2 Create the lead in the CRM, route it by region or size, and send a setup checklist before the kit arrives.
  3. 3 Watch for the first data from the trial device and confirm it is working, or offer help if nothing arrives after delivery.
  4. 4 Time sales follow-up to real usage, such as after a week of good data, with a short summary of what the device recorded.

Human oversight

Sales owns pricing, pilot terms and any promise about results. Trial data is shared with the customer only in the form they agreed to.

What gets easier

More trials end with a working installation, and sales conversations start from real data instead of guesses.

How it happens today

Dashboards exist, but nobody has time to open them. Out-of-hours heating or lighting goes unnoticed for weeks.

Systems and information

  • · Building management system, smart meters or IoT platform
  • · Spreadsheet or database
  • · Email

How it could flow

  1. 1 Pull last week's energy, occupancy and temperature data for each site.
  2. 2 Compare with the previous weeks and highlight unusual use, such as heating running in empty buildings or a meter that stopped reporting.
  3. 3 Write a short summary per site with the three things most worth checking.
  4. 4 Email it to the facility manager and, where wanted, to the owner or tenant.

Human oversight

Facility managers decide on any change to building controls or schedules. The summary points out patterns; it does not change setpoints.

What gets easier

Waste is spotted in a five-minute read each week, without anyone digging through dashboards.

From one task to a connected system

From a sensor reading to a resolved job

Linked together, these steps carry a signal from the device to a fix, a customer update and a bill, with people making the calls that need expertise.
  1. 01 Automated

    Reading received

    Devices report through a gateway, network and broker into the platform.

  2. 02 Automated

    Rules applied

    Thresholds, trends and heartbeats separate real events from noise.

  3. 03 Automated

    Incident opened

    A ticket or work order is created with readings and history attached.

  4. 04 A person reviews

    Technician assigned

    The on-call person accepts the job or it escalates to the next person.

  5. 05 People decide

    Fix on site

    A qualified technician diagnoses, repairs and records what was done.

  6. 06 A person reviews

    Customer updated

    The customer hears what happened, from approved templates.

  7. 07 A person reviews

    Usage and billing

    Usage is totalled and invoices drafted for finance to approve.

Responsible by design

What gets automated, what stays with people.

Can be automated

  • Collecting readings and checking heartbeats
  • Grouping duplicate alerts into one incident
  • Opening tickets and work orders with data attached
  • Weekly energy, uptime and excursion summaries
  • Setup checklists and first-data confirmations

AI prepares, a person checks

  • Threshold, escalation and on-call rules
  • Predictive maintenance suggestions
  • RMA approvals and replacement orders
  • Draft invoices and unusual usage flags
  • Firmware release notes and security notices

Stays with qualified people

  • Anything involving people's safety, fire, gas or electrical danger
  • Whether goods are fit for use after an excursion
  • Device design, firmware releases and rollout decisions
  • Vulnerability disclosure timing and regulator reporting
  • Pricing, refunds, warranty exceptions and contract terms

IoT rules differ by market and are changing. In the EU, the Cyber Resilience Act adds vulnerability reporting duties from 11 September 2026 and applies its main requirements from 11 December 2027, and the Data Act gives users access rights to connected product data. The UK PSTI regime has applied to consumer connectable products since 29 April 2024. The US Cyber Trust Mark is a voluntary FCC label still being set up. Sector rules such as US food traceability add their own records. Device data can be personal data under laws such as GDPR and US state privacy laws, so plan access, retention and security with care, and check current requirements with official sources or a qualified adviser.

What this gives people back

Less managing the process. More doing the work.

Alerts people trust

Fewer, better alerts mean a real problem gets attention instead of being lost in noise.

Technicians who arrive prepared

Every job carries the device, location, readings and history, so the first visit can be the last.

Customers who feel looked after

Offline devices, excursions and updates are communicated before customers have to ask.

Data that pays its way

Sensor readings turn into maintenance plans, accurate invoices and records you can produce when asked.

Built around how you work

Your process first. Technology second.

A startup with a few hundred devices on one vendor cloud and a facility group running meters and sensors across dozens of buildings need very different set-ups. What fits depends on your devices and connectivity, the platform or broker your data lands in, whether it offers MQTT, webhooks or an API, the help desk, CRM and accounting tools you already use and how your team handles on-call work.

Make It Flow begins by following a handful of real signals from device to resolution: an alert, an offline device, a return, an invoice. We note where data is copied by hand or lost. We then check what your existing tools can connect to, design small IoT automation workflows with clear approval points, test them against real readings with your team, and refine them as your fleet grows. Safety, engineering and every customer promise stay with your qualified people.

  1. 01Map the work as it really happens
  2. 02Find the steps worth automating
  3. 03Design with checks and approvals
  4. 04Connect, measure and improve
Need it built?IoT Development: device dashboards, sensor pipelines and alerts

FAQ

Common questions from iot and connected device businesses.

Still have a question?

Talk to us

IoT automation means acting on device data without someone watching a screen: a reading crosses a threshold and a ticket is opened, a device goes quiet and the customer is asked to check power, or usage is totalled and an invoice drafted. People still approve anything that affects safety, money or customers.

Often, yes. n8n has MQTT nodes for subscribing and publishing, and it can receive webhooks or call the REST APIs that most IoT platforms and vendor clouds provide. We check what your platform and plan actually expose before designing anything.

No. Workflows sit alongside what you have and connect it to your help desk, CRM, messaging and accounting tools. Dashboards stay useful for engineers; automation handles the follow-up work they do not cover.

It can be, if you start simply. Trend rules on vibration, temperature or run hours often catch problems early, and a small machine learning model is worth adding once you have enough history and recorded outcomes to check it against. We recommend proving value on a few machines first.

They mainly affect device makers: secure defaults, vulnerability handling, security updates and, in the EU, reporting duties. Automation can help keep records, track update notices and route vulnerability reports, but your engineers and advisers decide what compliance means for your products.

It should. We give each workflow only the access it needs, prefer self-hosted n8n where data must stay under your control, and keep personal data out of AI steps unless there is a clear reason and a lawful basis. Data protection duties depend on where your customers are.

Services

What we build for iot & connected devices.

Focused projects with fixed quotes, a written plan and a handover you own.
All services
Case Study Case Study: IoT Cold-Chain Monitoring & Support Automation How we built it

Free tools

Tools picked for iot & connected devices.

Free, quick and mostly sign-up free. Each one opens on the main Make It Flow site.
All free tools for this industry
Website Down Checker Check from outside whether your device API, customer portal or status page is reachable, so you know if a reported outage is on your side or the customer's network. SSL Certificate Checker Confirm the certificate on your API and portal hosts is valid, trusted and not close to expiry. Devices often fail hard on a bad certificate rather than showing a warning... TLS Version Checker See which TLS versions your endpoints accept, so you can retire old ones without cutting off older devices you still support. SSL Certificate & CSR Decoder Read the fields of a certificate you plan to pin or bundle in firmware, including expiry dates and subject names, before it ships. Security Headers Checker Grade the security headers on your customer dashboard and portal, the parts of a connected product that customers log in to. DNS Lookup & Health Check Check the records behind your API, broker and provisioning hostnames after a migration or a change of cloud provider. JSON Formatter & Validator Make a raw telemetry payload or API response readable and find the syntax error that broke a parser. CSV to JSON Converter Turn a CSV export from a data logger or sensor platform into JSON for an import, a test or a quick script. Base64 Encode & Decode Decode Base64 payloads that network servers often use for raw device messages, or encode test payloads for a downlink. Unix Timestamp Converter Convert Unix timestamps in device logs to readable dates, and back, when lining up readings with an incident. JWT Decoder Inspect the claims and expiry of a token your app or device uses to call the API, without sending it anywhere. Cron Expression Generator Build and check the schedule for heartbeat checks, nightly reports or batch uploads before you deploy it. Subnet Calculator Plan IP ranges for gateways and devices on a site network or VPN, and check whether two ranges overlap. Privacy Policy Generator Draft a starting privacy policy that explains what device and app data you collect, then have it reviewed for the markets you sell in. Email Deliverability Checker Make sure SPF, DKIM and DMARC are in place so alert emails, offline notices and RMA replies reach the inbox. QR Code Generator Print a static QR code on devices, packaging or install guides that opens your setup page or support form.

Free checklist (PDF)

Free iot & connected devices website checklist

29 practical checks for iot & connected devices: website speed and mobile, Google Business Profile and reviews, email security, booking and enquiry handling, and AI search visibility. Each one says why it matters, and many link to a free tool that checks it.

We email you the PDF and nothing else unless you tick the box. See the privacy policy.

Guides

Latest guides for iot & connected devices.

All guides

Connected devices deserve a connected process

Customers choose a device or a service because they trust it to work and trust you to notice when it does not. When the work behind every reading flows, your team can spend its time fixing, improving and selling rather than watching dashboards.

Usually a reply within one business day. No sales sequence, no obligation.