Case Study: Industrial Monitoring with IIoT, MQTT, and OEE

Industrial monitoring has evolved dramatically in the last decade. Traditional SCADA systems, wired PLC networks, and legacy monitoring stacks are being replaced by lightweight, scalable IoT infrastructures built on MQTT and BLE. This article is a detailed cluster page supporting our main MQTT Dashboard guide.

In this case study, we explore how a mid-sized manufacturing company transformed its operational visibility using a hybrid BLE sensor + MQTT broker + cloud dashboard architecture, powered by MQTTfy Dashboard.

What This Case Study Covers

Business problem
Technical architecture
Deployment challenges
Security strategy
Data modeling
Real-time visualization
Alert automation
ROI metrics
Scalability blueprint
Lessons learned

The Industrial Problem

Company Profile

Industry: Steel Component Manufacturing
Machines: 140 CNC machines
Location: Tier-2 industrial zone
Workforce: 220 employees
Operational Hours: 24/7

Key Challenges

No real-time machine health visibility
Reactive maintenance model
Unexpected downtime costing ₹3–5 lakhs/month
No centralized temperature & vibration monitoring
Manual data logging in Excel
Zero predictive insights

Existing System Limitations

Manual readings
PLC logs stored locally
No cloud aggregation
No remote monitoring

Required Capabilities

Real-time alerts
Historical analytics
Mobile dashboard
Low-cost wireless sensor layer
Scalable architecture

Why MQTT + BLE Was Chosen

Why Not Traditional SCADA?

Expensive licensing
Vendor lock-in
Hard to scale
Complex hardware

Why MQTT?

MQTT is a lightweight publish-subscribe messaging protocol designed for unreliable networks and low-bandwidth devices.

Advantages

Low overhead
Event-driven
Real-time communication
Supports QoS levels
Ideal for IoT environments

Why BLE?

Bluetooth Low Energy provides:

Low power consumption
Battery-powered sensors
Low-cost deployment
No heavy wiring
Easy retrofitting in existing factories

Final Data Flow Architecture

BLE Sensors → Gateway → MQTT Broker → MQTTfy Dashboard

Architecture Overview

graph TD subgraph "Level 0 and 1 Physical Process and Control OT" subgraph "CNC Machines" M1("CNC 1") M2("CNC 2") M3("...") M140("CNC 140") end subgraph "BLE Sensors" S1("Temp Vibration Sensor") S2("Current Sensor") end S1 -- "Attaches to" --> M1 S2 -- "Attaches to" --> M1 end subgraph "Level 2 Supervisory Control OT IT Bridge" Gateway1("BLE MQTT Gateway 1") Gateway2("...") Gateway12("BLE MQTT Gateway 12") Broker(("MQTT Broker")) S1 -- "BLE Adv" --> Gateway1 S2 -- "BLE Adv" --> Gateway1 Gateway1 -- "MQTT" --> Broker Gateway2 -- "MQTT" --> Broker Gateway12 -- "MQTT" --> Broker end subgraph "Level 3 and 4 Operations and Enterprise IT" Dashboard["MQTTfy Dashboard"] Historian["Time Series Historian InfluxDB"] Broker -- "Real time Data" --> Dashboard Broker -- "Data Stream" --> Historian end style Dashboard fill:#8B5CF6,stroke:#FFF,stroke-width:2px,color:#fff

Layer 1 – Edge Sensors

Installed on:

Spindle motors
Bearing housings
Compressor units
Control panels

Sensors Included:

Temperature
Vibration
Humidity
Current consumption (a key part of energy management)

BLE devices broadcast telemetry every 10 seconds.

Layer 2 – BLE Gateway

Industrial Raspberry Pi–based gateways, similar to those in our camera streaming tutorial, were used to:

Scan BLE advertisements
Parse sensor payload
Convert data into structured JSON
Publish to MQTT Broker

Topic Structure Used factory1/line3/cnc12/temperature factory1/line3/cnc12/vibration

Layer 3 – MQTT Broker

Managed MQTT broker cluster configured with:

TLS encryption
Device authentication
Topic-based ACL control
Persistent sessions

Layer 4 – MQTTfy Dashboard

All telemetry visualized using our MQTTfy Dashboard with:

Real-time graphs
Machine status indicators
Historical trend comparison
Alert triggers
Downtime heatmaps

The dashboard serves as the main visualization layer.

Deployment Strategy

Phase 1 – Pilot (15 Machines)

Goals:
- Validate BLE signal stability
- Test data loss rate
- Validate alert latency
- Check battery life
Results:
- 99.2% data reliability
- <1.5 seconds average alert delay
- 8-month projected battery life

Phase 2 – Full Rollout

140 machines
12 gateways
560 sensors
1 centralized MQTT cluster

Deployment completed in 28 days.

Data Modeling Strategy

One of the biggest mistakes in industrial IoT is poor topic hierarchy.

Structured MQTT Topic Format company/site/zone/machine/sensor

Example steelco/site1/zoneB/cnc14/vibration

Benefits

Scalable filtering
Easy wildcard subscriptions
Clean multi-tenant expansion
Simplified dashboard mapping

Payload Example

{
  "timestamp": 1707001122,
  "value": 48.3,
  "unit": "celsius",
  "battery": 87
}

Alerting & Automation

Threshold-Based Alerts

Trigger Conditions:
- Temperature > 75°C
- Vibration > 20 mm/s
- Current spike > 15%
Trigger Actions:
- Email notifications
- WhatsApp alerts
- Dashboard notifications
- Machine shutdown (future integration)

Average detection time: 1.2 seconds

Security Implementation

Industrial systems are highly sensitive.

Security Layer	Implementation
Transport Security	TLS encryption, Broker certificate validation
Authentication	Unique client ID per gateway, Rotating authentication tokens
Access Control	Topic-level ACL rules, Read-only dashboards for supervisors
Network Segmentation	VLAN separation, Firewall rules for broker

No security breaches recorded post-deployment.

Downtime Reduction Results

Metric	Before Implementation	After 6 Months
Unexpected breakdowns/month	11	3
Monthly downtime losses	₹4.2L	₹1.1L

ROI

System Cost: ₹28L
Average Monthly Savings: ₹3L
ROI Achieved In: 9.3 months

Predictive Maintenance Model

Using historical vibration trends and effective data visualization, we achieved predictive maintenance, a technique also vital in our fleet management case study.

Bearing failure predicted 9 days early
Compressor overheating predicted 5 days early

Techniques Used

Rolling average anomaly detection
Baseline deviation alerts
Historical trend visualization

Future Plan: ML-based predictive modeling integration.

Dashboard Capabilities (MQTTfy Integration)

Following the principles of building an effective IoT dashboard, we used MQTTfy Dashboard for:

Live sensor tiles
Multi-line graphs
Time-series export
Machine grouping
Mobile-responsive interface
Role-based access
Historical comparison
CSV export
Industrial dark theme

Dashboard Functions

NOC monitoring center
Maintenance planning tool
Executive reporting engine

Scalability Plan

Next Phase

Multi-factory deployment
1200+ machines
Cloud auto-scaling broker
Geo-distributed gateways

This level of scaling is similar to the requirements for a smart city project.

Designed Capacity

100K+ MQTT messages per minute
Horizontal broker scaling
3-year data retention

Lessons Learned

Start with structured topic naming.
Run a pilot before full rollout.
Monitor gateway CPU & memory.
Enforce strict ACL rules.
Track sensor battery levels.
Design dashboards around business users.

Why This Case Study Matters

Industrial monitoring is no longer optional. It's as crucial for manufacturing as it is for retail analytics.

Factories adopting MQTT + BLE gain:

Real-time operational visibility
Reduced downtime
Predictive maintenance capability
Centralized control
Lower wiring costs
Seamless scalability

Legacy SCADA systems cannot match the flexibility of modern IoT-first stacks.

Conclusion

This industrial monitoring deployment demonstrates how a hybrid MQTT + BLE architecture can:

Reduce downtime by 70%
Achieve ROI in under 12 months
Scale from pilot to multi-factory
Enable predictive intelligence
Replace expensive SCADA systems

The MQTTfy Dashboard acts as the visualization and intelligence layer, transforming raw telemetry into actionable industrial insights for everything from factories to smart homes.

Industrial IoT is no longer future-ready — it is present-critical.

Frequently Asked Questions

What is the difference between IT and OT?

IT (Information Technology) deals with the systems that manage data, like servers, databases, and business applications. OT (Operational Technology) deals with the systems that control physical processes, like PLCs, SCADA systems, and industrial machinery. IIoT and MQTT are critical for bridging the gap between these two worlds.

How do you get data from a PLC into an MQTT dashboard?

You use an MQTT Gateway or an IIoT platform. This software or hardware device connects to the PLC using its native protocol (like Modbus, OPC-UA, or Siemens S7), reads the desired data registers, converts the data into a structured format like JSON, and then publishes it to an MQTT broker. The dashboard then subscribes to these MQTT topics.

What is OEE and how is it calculated?

OEE (Overall Equipment Effectiveness) is a key performance metric for manufacturing that. It is calculated as the product of three factors: Availability (runtime / planned production time), Performance (actual cycle time / ideal cycle time), and Quality (good parts / total parts).

What is Sparkplug B and how does it relate to MQTT?

Sparkplug B is an open-source specification that defines a standard topic structure and payload format for sending IIoT data over MQTT. It provides a plug-and-play framework that ensures interoperability between different industrial devices and. It adds much-needed context (data types, historical state) to raw MQTT messages.

How does MQTT enable predictive maintenance?

By continuously collecting and transmitting high-resolution sensor data (like vibration, temperature, and current draw) over MQTT, you can feed this data into machine learning models. These models can learn the normal operating signature of a machine and detect subtle anomalies that are precursors to failure. This allows you to schedule maintenance before a breakdown occurs, saving significant time and money.