Smart manufacturing with MQTT: OEE, predictive maintenance, energy monitoring and real-time production data flowing through a managed MQTT broker.
A clear walkthrough of how to calculate OEE: the three factors, a full worked shift example, the mistakes that inflate the number, and Python that computes OEE from machine-state MQTT messages.
Condition monitoring tells you a machine is unhealthy now; predictive maintenance estimates when it will fail. Learn the difference, the maturity path between them and how the MQTT data flow changes at each step.
OEE, OOE and TEEP share the same performance and quality factors but measure against different time bases. This comparison works through one example and lists the MQTT signals each metric needs.
Tablets, smartwatches and line displays become far more useful when they receive live events. Learn connected worker use cases, the MQTT topic design and a browser client that receives tasks over secure WebSocket.
Build a working vibration monitoring sensor from an ESP32 and an accelerometer, compute RMS, peak and crest factor on the device, publish them over MQTT and raise alerts when a machine drifts from its baseline.
Packaging lines lose most of their time in short stops nobody records. This use case shows how to capture every stop and speed change over MQTT and turn them into a micro-stop Pareto.
ERP plans, MES executes and SCADA supervises. Learn what each system owns, where they overlap, how they map to ISA-95 levels and how an MQTT broker links them without point-to-point spaghetti.
Monitor freezers and cold rooms with ESP32 sensors publishing over MQTT, detect real temperature excursions without nuisance alarms from door openings, flag offline sensors and keep an excursion log.
ISO 50001 is a management standard, but it runs on data. This explainer maps the standard's data requirements to meters, relevant variables and an MQTT-based collection architecture you can audit.
A hands-on tutorial for real-time OEE monitoring: machines publish states and counts over MQTT, a Python service calculates availability, performance and quality every minute and publishes live OEE for any dashboard.
Track downtime on any machine, old or new: an ESP32 with a clamp-on current sensor publishes run and stop states over MQTT, and a Python logger records every stop with its duration and reason.
A practical blueprint for an MQTT-based andon: how operators raise calls, how call states are modelled with retained messages, how escalation works by time, and how response-time breaches become alerts.
A practical bearing temperature monitoring setup: battery-powered wireless sensors on bearing housings, ambient-compensated readings over MQTT, rate-of-rise detection and threshold alerts that fire before a bearing overheats.
Learn how to collect execution state, active program, spindle load and part count from CNC machines, publish them over MQTT and turn them into an honest utilization number.
A practical use case for injection molding monitoring: publish one MQTT message per shot with cycle time, cushion, pressure and temperatures, then flag drifting shots before bad parts pile up.
Manual sheets, PLC tags, retrofit sensors and edge gateways all collect shop floor data differently. Compare the methods, pick the right one per machine and centralize everything on one MQTT namespace.
Learn how in-line sensors, gauges and test stations publish measurements over MQTT for real-time quality checks, how to design topics and payloads, and how the same data feeds SPC and traceability.
What predictive maintenance with IoT really takes: the sensors that matter, features to compute at the edge, how condition data moves over MQTT to analytics, and the modelling steps from thresholds to anomaly detection.
Count good and rejected parts automatically with photoelectric sensors and a Raspberry Pi, persist counts across restarts and publish retained shift totals over MQTT so every screen shows the right number.
Connect DIN-rail Modbus energy meters on an RS-485 bus to a gateway, decode the registers in Python and publish per-machine kWh to an MQTT broker so every machine gets its own energy figure.
A practical blueprint for plant-wide energy monitoring: which meters to install, how to structure MQTT topics, how to store counters and how to report energy per line, per shift and per part.
A step-by-step build of a production line monitoring dashboard: stations publish status and counts over MQTT, Telegraf stores them, and Grafana or Node-RED turns them into a live line view.
Stream measurements over MQTT, compute X-bar and R control limits in Python, apply Western Electric rules to every new subgroup and show out-of-control points on a live dashboard within seconds.
Lean teaches you to see waste; IoT data lets you see it continuously. Learn how real-time machine events support value stream mapping, kaizen, SMED, andon and TPM without replacing gemba.
Build a working batch process monitoring pipeline with MQTT. Publish batch phases and process values, compare them to recipe setpoints and record deviations in real time with Python.
Measure every single cycle instead of sampling with a stopwatch. Capture cycle signals with an ESP32, publish them over MQTT and compute rolling statistics that reveal the real bottleneck of a line.
Define a clean machine state model, derive running, idle and down from stack lights or PLC signals, and publish retained MQTT messages so every dashboard shows the current state the moment it connects.
Replace brittle point-to-point MES interfaces with an event-driven MQTT namespace. Learn the topic layout, payload contracts and a working adapter for orders, counts and quality results.
How to build a downtime reason code tree that stays short, specific and actionable, plus an operator input flow that publishes reasons over MQTT alongside automatically detected machine states.
Food and beverage plants live on temperature, hygiene and throughput. See the IoT use cases that matter most, how to structure the MQTT data and what to watch out for in wet, regulated environments.
Create an account, get your broker credentials and connect your first device — no servers to install, patch or scale.
