There are now more than 21 billion connected IoT devices worldwide, with the market set to nearly double in the next five years.
To unlock the full potential of connected devices, clear and consistent communication is key. That’s where message queueing telemetry transport (MQTT) comes in—a reliable, lightweight messaging protocol built on a publish/subscribe framework to enable real-time communication across industrial networks.
Designed for low-bandwidth, low-power, high-latency environments, MQTT is ideal for factory floor and remote assets and is central to scaling Industry 4.0 initiatives such as real-time reliability monitoring.
Read on to learn more about the MQTT essentials, the purpose of the protocol and how it helps set the stage for manufacturing success.
What is MQTT? Definition and core purpose
MQTT was developed in 1999 to facilitate satellite-based oil pipeline monitoring. To account for both the bandwidth limitations of late 20th-century networks and the geographical scale of pipeline operations, MQTT was designed to use minimal bandwidth and power.
This makes it ideal for managing Industrial IoT (IIoT) devices in manufacturing environments by enabling the fast and reliable transmission of real-time equipment data, even when network conditions are less than ideal.
Devices that leverage MQTT in manufacturing include connected predictive maintenance sensors, such as those used to measure temperature, pressure or vibration, along with programmable logic controllers (PLCs) and production line machinery.
How MQTT works—the publish/subscribe model explained
MQTT uses a publish/subscribe architecture to facilitate the exchange of data. This differs from a traditional client/server framework, also known as request/response.
In a request/response relationship, clients query servers when they need data. In a publish/subscribe model, clients send a single request for ongoing information, after which the server provides this data at fixed intervals or in response to changing conditions.
Consider a critical piece of production line equipment that experiences significant temperature variations as part of its standard operational process. If these variations exceed a specific threshold, however, the result may be irrevocable damage to the machine. Maintenance teams install an IIoT sensor to record temperature changes and use a computerized maintenance management system (CMMS) to track these changes and take action as needed.
In a client/server approach, CMMS tools must actively query sensors, either using a fixed schedule or in response to user actions. In a publish/subscribe model, the sensor publishes temperature data at regular intervals, and this data is captured by CMMS for predictive maintenance. The publish/subscribe model has three key components:
- Publisher: The publisher in MQTT is the device that transmits data. Data is published as a topic accessible to any subscriber, rather than sent as a direct message to specific endpoints.
- Subscriber: The subscriber is the device that receives the MQTT topic information. This could be a backend CMMS, EAM or ERP system, or it could be an end-user mobile device. Subscriber devices can sign up for multiple topics.
- Broker: The MQTT broker facilitates communication between publishers and subscribers. It receives requests for topic subscription and sends the topic data to all subscribed devices. This ensures that the publishing device only has to push new data once—after that, the broker handles distribution. Both publishers and subscribers are known as MQTT clients.
Because devices communicate via topics rather than direct connections, publish/subscribe frameworks are more scalable than point-to-point client/server solutions. They also offer lower network load and help decouple data producers and consumers—subscribers can subscribe to multiple publishers and receive updates simultaneously.
MQTT vs. traditional industrial protocols
MQTT isn’t the only industrial protocol used for IIoT applications. Other common options include OPC UA, Modbus, BACnet and HTTP APIs.
For example, OPC UA delivers more data-rich insights than MQTT, but requires more bandwidth. Modbus is simpler but typically works best with legacy devices, while BACnet is specifically designed for building automation and network control. HTTP APIs are useful for bridging the gap between disparate tools and technologies but rely on a traditional client/server model.
As a rule, MQTT outperforms its protocol counterparts in multi-site, multi-vendor IIoT ecosystems, thanks to benefits such as:
- Greater speed and efficiency: MQTT is lighter, faster and more efficient than polling-based protocols. The protocol sends messages using the smallest amount of data possible and only needs to send data to a single location—the broker.
- Push-based communication: MQTT uses push-based communication, while traditional protocols often use pull-based methods. In practice, this means MQTT “pushes” the data to brokers as needed, and this data is then sent to all subscribers. Other protocols require users to request or “pull” this data directly from sensors or other devices.
- Lower bandwidth consumption: Smaller MQTT packet sizes mean reduced data transmission volumes and, in turn, lower bandwidth consumption. This allows MQTT to operate effectively even on slow or unstable networks.
- Ease of scalability: Because each publisher must connect to only a single broker, and these brokers can be used to handle data from multiple publishers, MQTT environments are easily scalable.
It’s worth noting that MQTT does not replace other standards. Instead, it works alongside different protocols to improve visibility and deliver consistent manufacturing performance.
Why MQTT matters for modern manufacturing and maintenance
Adopting MQTT offers several benefits for modern manufacturing.
First is continuous condition monitoring across critical assets—push-based MQTT communication ensures that teams always have access to current and accurate machine data. Next is support for predictive maintenance plans. While preventive maintenance can help reduce the risk of equipment failure, the real-time sensor data provided by publish/subscribe models gives teams the insight they need to proactively address possible issues.
In addition, MQTT reduces latency, which allows faster anomaly detection and supports a lower mean time to detect (MTTD). The MQTT protocol also improves data access across CMMS, enterprise asset management (EAM) solutions, machine health monitoring systems, visual dashboards and AI models.
The result? Companies are better equipped to unlock true factory-connected capabilities.
MQTT quality of service (QoS) levels explained
MQTT has three QoS levels. Here’s what each means and how it can impact performance.
- QoS 0: At most once: This is the fastest, least resource-intensive QoS. As the name states, information is only delivered once, and delivery is not guaranteed. Consider a temperature sensor that detects an anomaly and pushes this data to a broker. Once the data is sent, it will not be sent again even if the MQTT message is lost, the data is corrupted or the broker does not receive the message.
- QoS 1: At least once: In Qo2 level 1, messages must be delivered (and received) at least once. This means that no message will be lost, but retained messages may be sent and received multiple times.
- QoS 2: Exactly once: Qo2 is the most reliable option but also requires more overhead. In this case, message delivery happens exactly once. They are not sent again but are guaranteed to reach the broker.
In most preventive maintenance use cases, MQTT QoS 1 offers the best balance between reliability and resource use.
MQTT in Industry 4.0, smart factories and predictive maintenance
As companies move to deploy Industry 4.0 best practices and build automated factories supported by predictive maintenance and digital twins, MQTT integration becomes essential. MQTT easily integrates with:
- Edge computing gateways
- Cloud analytics platforms
- PLCs, SCADA, CMMS and EAM systems
In practice, MQTT powers solutions such as real-time asset dashboards, AI-driven anomaly detection and automated maintenance triggers, all of which are essential for companies to manage multi-site maintenance and support centralized monitoring strategies.
Common MQTT use cases in manufacturing and maintenance
The lightweight nature of MQTT makes it ideal for multiple use cases, such as:
- Condition monitoring
- Sensor data transmission
- Automated maintenance notifications
- Energy monitoring
- Remote asset management
- OEM equipment reporting
Mapping MQTT to maintenance and manufacturing processes
Manufacturing and maintenance programs are evolving. For example, the International Society of Automation (ISA) suggests that even as Industry 4.0 adoption faces challenges in widespread adoption, there’s a push toward Industry 5.0, which focuses on the “human touch”—the combination of human expertise and collaborative robots to deliver customized and highly specialized outputs.
These IIoT trends create the need for reliable, real-time communication across connected device networks. This is the role of MQTT: enabling smarter, faster and more reliable maintenance operations while also providing the visibility necessary for companies to identify predictive maintenance opportunities and performance improvement pathways.
Discover how ATS can help your company implement end-to-end MQTT-based systems to improve system uptime and drive real-time insight. Let’s talk.
References
Amazon Web Services. (n.d.). What is MQTT? AWS. Retrieved December 22, 2025, from https://aws.amazon.com/what-is/mqtt/
Sinha, S. (2025, October 28). Number of connected IoT devices growing 14% to 21.1 billion. IoT Analytics. Retrieved December 22, 2025, from https://iot-analytics.com/number-connected-iot-devices/
Østergaard, E. H. (n.d.). Welcome to Industry 5.0. InTech. International Society of Automation. https://www.isa.org/intech-home/2018/march-april/features/welcome-to-industry-5-0
