Connectivity turns sensor signals into data that enables machine condition monitoring and provides insights into ways of improving operational performance. If advances like predictive maintenance form part of your planned digital transformation, understanding the underlying principles of this technology will help you move forward.
Household devices like doorbells, baby monitors and refrigerators become more useful when they are connected to the internet, and the same is true for industrial sensors. While sensors were once limited to functioning within machine control systems, internet technologies enable them to feed data into centralized platforms.
These platforms not only store and analyze data but also aggregate data from multiple devices. This creates opportunities to see patterns and trends at a larger scale, revealing insights into ways of preventing problems and making improvements.
Hooking devices like sensors (and doorbells and baby monitors) onto the internet creates an Internet of Things (IoT). This blog explains, from the perspective of industrial applications, what IoT sensors are, how they work and how an IoT sensor differs from a smart sensor. It reviews the use of IoT sensors in maintenance, covers the basic elements of an IoT solution for condition monitoring and explores the biggest implementation challenges. If you’re ready to learn more about industrial IoT sensor technology, read on.
What are IoT sensors?
Sensors measure physical properties and output a value. Some of the properties most often measured are temperature, pressure, humidity, acceleration and electrical current and voltage. Recent advances in materials and solid state electronics have miniaturized these sensors to where most can be integrated into small printed circuit boards.
Temperature sensors, for example, often use Negative Temperature Coefficient (NTC) thermistors while piezo-based sensors measure pressure and acceleration. (Vibration measurement starts out as a measurement of acceleration, with the data being processed for frequency and amplitude.)
In very simple systems, sensor outputs are observed by an operator or engineer. If the output starts to drift outside the target range, they will make an adjustment to the input conditions to bring it back to the desired level.
In sophisticated modern control systems, sensor outputs are used as inputs to the control system with no human involvement required. PID feedback control is perhaps the most obvious example.
A limitation of basic sensors is that their output signals can often be noisy. To address this, they may have signal conditioning capabilities, such as amplification and filtering, built-in. They may also incorporate analog-to-digital converters, pre-processing and decision-making logic, and format output values for a communication protocol such as USB or OPC-UA.
An IoT sensor is similar to a smart sensor, but with added computation and output capabilities. In addition to the sensing elements, it typically consists of a microcontroller, a power management module, a communications module and possibly also some data storage.
How do IoT sensors work?
In an IoT sensor, the process from data collection to actionable insight is:
- Sensing: Current and/or voltage are produced in response to measured physical conditions.
- Signal conditioning: The signal is amplified, filtered, and possibly subjected to other processing in order to improve its clarity and value.
- Signal digitizing: An analog-to-digital converter turns the signal into a form that can readily be processed and stored electronically.
- Signal processing: To minimize the amount of data transmission needed, the signal may be treated in a way that reduces the quantity of information to be transmitted. This might include averaging over a period or identifying when limits have been exceeded. When complex processing is needed, the IoT device is often said to incorporate edge computing capabilities.
- Transmitting: Using the communications module, the IoT device sends data to a gateway device. Like a router, this can forward the data on to other devices. This could be an IoT platform running on a server or PC or a cloud server where the data is stored and/or analyzed and processed.
- Storing: Data may be stored on the device (essential if it is only transmitted periodically), or the gateway can hold it or send it to another device, such as a cloud server, for storage.
- Analysis: The data is retrieved and analyzed to generate the required insights.
Power management is an important aspect of IoT sensors. In industrial settings, sensors can often receive their power either by dedicated 24V supply or through the communications interface. (USB and Power over Ethernet [PoE] do this.) Some though will need their own internal power source, either due to distance from the gateway or the type of installation location. (An example is when a sensor is mounted on rotating or other mobile equipment.)
Usually, internal power is provided by a long-duration battery, although energy harvesting technologies are emerging. These power the sensor from sources such as the energy recovered from machine vibrations or radio transmissions.
Two key considerations in power management are:
- Will the sensor be “always-on”?
- The frequency and volume of data transmission
The most efficient strategy is to sample the physical property intermittently and summarize the data for transmission, possibly keeping it as simple as “over/under limit.” Whether this is feasible depends on the type of data being collected and whether it is needed in real-time, only for trend analysis at some later time, or some intermediate requirement.
Communication protocols used by IoT sensors
When deciding how an IoT sensor should transmit its data, the main considerations are:
- Reliability: Despite the proliferation of wireless protocols, many plants still prefer serial communication, wired Ethernet or USB. Cable installation can be costly and time-consuming, but these provide reliable connectivity with good immunity to the electrical noise sources often present in busy plants (arc welding, for example).
- Distance/range: While wired Ethernet has much better range than USB— which needs repeaters/amplifiers after relatively short distances—cable runs on large sites can become very expensive. Wireless technologies are therefore often preferred for remote monitoring applications.
- Power consumption: For IoT devices running on batteries, it is essential to minimize power consumption. This promotes the use of lower-power communications protocols.
- Compatibility with receiving hardware: Ethernet is ubiquitous in most manufacturing plants, most newer machines also use USB, and many will work with OPC-UA, the standards for exchanging data between industrial equipment. In addition, serial communication (RS232, RS485), an older technology, still has a role in many automation and process control setups.
When looking at wireless protocols, after Wi-Fi (used in most plants) there are Zigbee and Bluetooth for short-range communications. For long range wireless communication, the main options are narrowband IoT (NB-IoT) which uses the LTE cellular spectrum, and Long Range Wide Area Network (LoRaWAN) operating in the unlicensed part of the radio spectrum.
Given the challenges of power management and battery life, these long-range technologies are generally used only for periodic transmission of small data packets.
Applications of IoT sensors in industrial maintenance
IoT pressure sensors, flow sensors, temperature sensors and more are revolutionizing maintenance of manufacturing equipment. They enable machine condition monitoring—especially valuable for remote assets like wastewater treatment plants—and support predictive maintenance through data aggregation and analysis.
Industrial applications include:
- Monitoring electrical current draw as an indicator of cutting tool condition
- Measuring spindle vibration to detect bearing wear
- Monitoring particulate levels in fluids to detect contamination
- Tracking pump output pressures and flow rates for pump and filter monitoring
In each of these and in myriad other applications that exist, the IoT sensors can provide an alert prior to a problem emerging. This allows replacement parts to be ordered, and work scheduled, before a breakdown occurs.
Use cases go beyond these machinery monitoring scenarios though. Other industrial applications of IoT technology include environmental control (HVAC systems), asset tracking and safety systems.
Benefits of IoT sensors for manufacturers
The true value of IoT sensors lies in their ability to communicate data to systems with the ability to aggregate and analyze it. This provides new insights into equipment condition and operation, enabling extended lifespan while reducing maintenance costs and both planned and unplanned downtime.
More specifically, the benefits can be grouped under four headings:
- Increased equipment uptime, reduced unplanned downtime
- Energy savings and process optimization
- Data-driven decision-making and real-time visibility into operations
- Support for scalable asset management and predictive maintenance programs
Every temperature sensor, pressure sensor, humidity sensor, light sensor and motion sensor, once given computing and communications capabilities, can become an IoT-connected device. This provides the maintenance team with reams of previously unavailable data, allowing them to identify patterns and trends to optimize machine maintenance.
Precise, timely information on processes enables higher levels of optimization, improving product quality while simultaneously reducing energy consumption. For example, by reporting fluid levels, minute-by-minute if necessary, a connected ultrasonic sensor can warn when oil levels are low, preventing unplanned stoppages or breakdowns.
While connected devices can provide real-time warning of current or potential problems, advanced analytics programs are capable of aggregating and analyzing large volumes of data. This brings to the fore patterns and trends that would otherwise be lost or overlooked. Thus, IoT sensors support real-time visibility while providing new insights that support data-driven decision-making.
An IoT application should be scalable, rather than being tied to just a single piece of equipment. Choosing the right platform supports scalable asset management and enables the use of predictive maintenance programs.
Challenges and considerations when implementing IoT sensors
Whether dealing with a simple proximity sensor confirming a gate is closed or an autonomous mobile robot has docked, or an infrared sensor monitoring gas emissions, IoT sensors present some unique challenges.
At the level of the sensors themselves, these relate to accuracy and calibration, and for remote sensors, power consumption and battery life.
The starting point is sensor selection. Choose devices from reputable manufacturers with a track record of delivering reliability. Ensure their sensors arrive factory-calibrated, and then once installed, monitor for sensor drift or use redundant sensors to detect anomalies. Where possible, perform periodic remote calibration; alternatively, seek out self-calibrating smart sensors.
For remote, battery-powered sensors, power consumption is minimized by compressing data packet size and limiting transmission frequency. This requires careful analysis of the data requirements and the purpose for which it is being collected. (Real-time monitoring of remote assets may be better served by providing a reliable power source rather than relying on batteries.)
A major concern in all wireless data transmission systems is security. This is addressed by working closely with the sensor vendors and providers of services like cloud storage and predictive maintenance, and by making informed decisions about the communications protocols being employed.
When deciding to adopt IoT technology, this should form part of a digital transformation strategy rather than a standalone project. At a minimum, explore integration with SCADA systems, and look also at how readily sensors will connect to the CMMS. Ease of ERP integration (supported by OPC-UA) may also be a consideration, depending on the goals of the organization.
Finally, no organization wants to make a substantial investment in IoT technology, only to find either it quickly becomes obsolete, or support goes away, or it cannot be expanded as the business grows. To avoid such problems, consider following a two-part strategy that focuses on partner selection and education in the technology.
Risk is minimized in IoT implementation, as with other digital transformation technologies, by working with an experienced partner, so be sure to evaluate their previous work in this space. In parallel, learn more about the technology as this will help you ask better questions and define your requirements more carefully.
How outsourced maintenance supports IoT-driven strategies
ATS was established more than 40 years ago to help industrial plants with their maintenance challenges. Today, in addition to short-term support, assistance with parts sourcing and management, and fully outsourced maintenance, ATS offers a range of condition monitoring, IIoT integration and predictive maintenance services. Tailored to meet the needs of businesses across multiple sectors, these solutions offer the experience and expertise necessary for an effective digital transformation.
Effective maintenance is of critical importance in today’s challenging manufacturing environment, and connected maintenance can improve uptime. Contact ATS to learn more.
