Automation is crucial in industrial sectors. From manufacturing to food and beverage and power industries, automation is used to improve productivity, quality, efficiency and safety of the many processes in different industrial sectors. This includes monitoring production lines, ensuring the safety of power generation operations, including the assembly and packaging of products.
At the core of industrial automation are sensors. There are many different sensors used within industrial settings to measure different parameters in real time and ensure that the lines and processes continue to work smoothly. Alongside measuring environmental parameters, sensors also track the position and appearance of physical products to ensure they remain a high quality, and the process lines run optimally. Many different sensors are required across different sectors, and their accuracy, precision and positioning within are key to ensuring that industrial operations stay automated.
The role of sensors
Many different sensors are required for industrial processes that are automated due to the vast number of parameters that need to be measured. “All these sensors gather real-time data about various aspects of the industrial process, enabling informed decision-making and process control” says William Buratto, researcher at Universidade do Estado de Santa Catarina, Brazil.
Speaking to Electronics Weekly, Haroan Fu, research professor at China’s Zhejiang University says that “sensors for automated industrial systems are used to perform control, efficiency improvement, safety, and quality control (QC) functions”. On the control side, monitoring the industrial processes provides feedback to the control system that allows real time adjustments to be made. For efficiency, automated systems can perform tasks much quicker than humans to improve productivity, but this is only possible if sensors are there to provide the data that allows these tasks to be performed. In automated systems, the sensor data can measure hazardous systems that will trigger an alarm or shutdown process if a hazard is detected. In automated QC operations, sensors will detect defects by measuring the shape and size of products and alert the control system to take corrective action, ensure that products are manufactured to consistent and precise specifications, and analyse the data to identify trends that may indicate quality issues.
Some of the main sensors used to perform these industrial operations are detailed below.
Temperature sensors
Temperature sensors collect data in their local environment and convert it into a quantifiable output. Sensors that can measure the relative humidity of the local area are often used alongside temperature sensors.
Temperature sensors are used to monitor and control the temperature of different industrial processes, such as measuring the internal temperature of cleanrooms and server rooms, monitoring the process temperature in semiconductor fabrication lines to ensure high-quality wafers, and monitoring the temperature of industrial equipment – such as pumps, motors, chemical reactors, and food processing equipment.
Temperature sensors are also widely used in energy generation processes to provide better reliability and safety. Buratto says that “most temperature sensors in power plants are directly wired to the control systems and are used in conjunction with transmitters to save time and money during installation, improve measurement reliability, reduce maintenance, and increase uptime”.
Temperature sensors are critical in power plants because of the high temperatures that are generated during operation – especially during peak demand periods – as temperatures that are too low or too high can damage the critical (and expensive to replace) equipment. “The unforgiving industrial environments in power generation requires precision and accuracy in the harshest of high temperature conditions” says Buratto.
Examples of temperature sensors in power plant operations include measuring furnace temperatures in energy-from-waste plants, detecting faults and damage to turbine blades in gas turbine plants, and monitoring the operating temperature in coal-fired power plants so they can operate with a higher efficiency.
Pressure, flow and force sensors
Pressure, flow and force sensors have many uses in automated industrial processes. Pressure sensors convert a localised pressure into an electrical signal, with the value being relative to the pressure. Force sensors are similar but produce a signal under an applied force/load. Vacuum sensors are a specialist type of pressure sensor that are also used for measuring pressures below atmospheric pressures and use a heated wire to detect changes in temperature convection in the local environment. By correlating it with electrical resistance, the vacuum pressure can be deduced. Flow sensors are also specialist pressure sensors that calculate the pressure difference in a fluid or gas to determine its flow rate.
These sensors are widely used for analysing flow and pressure properties in industrial hydraulic and pneumatic systems – such as pumps and compressors – as well as for detecting leaks and blockages in process equipment. In the food processing space, they monitor the pressure of the steam in sterilising and pasteurising equipment. Force sensors also measure the forces required to insert individual components into products without breaking them and for testing the durability and strength of materials across industrial sectors.
In the power plant sector, Buratto states that “pressure sensors are used to measure steam pressure and flow rates in turbines, measure the flow of fuels to control combustion processes in different power plants, and determine the flow rates of cooling water flows to ensure that equipment is maintained within its optimal temperature range”.
Proximity and positional sensors
Motion and position sensors detect any objects that move or change their position and detect these changes in either a rotary or linear motion. The movement signals are processed, allowing a high degree of control of all objects in an automated industrial environment. Proximity sensors also detect objects around them by measuring a change in an induced magnetic field and can be used in water and oil process environments unlike positional sensors.
Linear and rotary position sensors are used to detect objects and components on robotic assembly lines and conveyor belts to provide a more precise control of the automated lines. They are also used to track the positions of valves and doors in process environments and identify the precise positions of machining tools.
Proximity sensors are also used to track the position of materials, parts and components on conveyor belts and monitor the position of different tools such as lathes, milling, and drilling machines. They are also used in process environments because they aren’t affected by oil or water contamination.
Vision sensors
Vision sensors convert an optical image into an electrical signal. Vision sensors play a key role in automated and smart manufacturing environments by identifying defects in the manufacturing line, tracking the movement of products and components, and ensuring that all products meet the intended quality standards.
There are two main types: charge coupled device (CCD) and complementary metal-oxide-semiconductor (CMOS). CCDs use a charge-coupling mechanism to convert an electrical charge from individual pixels to a readout circuit, where CMOS vision sensors are fabricated with a pixel and readout circuit on the same chip.
Finally, there are also level sensors. These are used to measure the levels of liquid and solids within process tanks, containers and silos, typically in the food and beverage and wastewater treatment industries. “Level sensors prevent the liquid level from being too high and overflowing, as well as ensuring that the levels within tanks does not drop too low” says Buratto. Keeping an eye on tank levels can also help to identify any potential leaks in different process environments.
Design considerations
There are many design considerations that need to be considered when fabricating sensors for industrial settings – to cope with both the demanding environments and the need for ensuring a high product quality (or a high safety if it’s a power generation process).
Fu says that “adapting to harsh environments, stability and durability, electromagnetic interference protection, and integration complexity are some of the main design challenges for sensors used in automated industrial processes”. Buratto confirms that “complexity, the need for expertise, and data security and confidentiality are three major design challenge areas”.
The two experts discussed the main design considerations for sensors in industrial automation with Electronics Weekly, and their thoughts are laid out in the following table:
| Design challenge | Design issues | Potential solutions |
| Harsh environment adaptation | Temperature, humidity, vibrations, and chemical corrosion can impact sensor performance. | Using materials that are resistant to high temperatures, mechanical stimulus and/or chemical corrosion help to improve industrial sensor performance. Applying protective encapsulation layers can also protect sensors in harsh operating conditions. |
| Stability/durability | The need to maintain a high accuracy and stability over long time periods without maintenance. | Components that are resistant to wear are key to longevity and automated self-calibration helps to prolong sensor usage. |
| EMI protection | EMI and noise can affect sensors in industrial automation systems. | Use shielding designs and noise filtering techniques to improve the sensor’s resistance to interference. |
| Complex integration | The need to balance cost with compatibility with existing equipment. | Modular design approaches can help with computability, scalability and cost considerations. Eliminating non-essential functions and selectively integrating key sensing features can help to reduce both costs and integration complexity. |
| Need for expertise | Integrating IoT sensors to monitor industrial process in real-time requires specific expertise. | Engineers with specific expertise can help to ensure that IoT sensors are efficiently connected to existing control systems, allowing for the precise transfer of data in real-time. |
| Data security issues | Data collected by IoT sensors needs to be secure and confidential. | Robust cybersecurity and data protection processes need to be in place at the time of integrating and constantly managed to ensure long-term reliability and safeguarding of data. |
Adding AI capabilities
Artificial intelligence (AI) algorithms, such as machine learning (ML) and deep learning, are now starting to improve sensor capabilities in automated process, as they can better analyse large data sets to optimise sensor performance and improve the accuracy of control systems that rely on this data. The integration of these algorithms is also enabling automated systems to be equipped predictive maintenance capabilities to reduce downtime. Fu states that AI/ML are “providing advanced data analysis, automated data cleaning, real-time monitoring, advanced data fusion, and smart control capabilities to automated industrial processes”.
Alongside the advanced capabilities in the industrial settings, AI/ML is also helping to improve the core sensor technology through “inverse design, response optimisation, and automatic calibration and adjustment” says Fu. “This leads to industrial sensors for automated processes that have an enhanced sensitivity and accuracy, shorter response time, a higher efficiency, higher fault prediction accuracies, and are more robust in complex and changing environments”.
The future of industrial automation
Sensors are becoming increasingly advanced, and more capabilities are being added to them to improve the robustness and accuracy of industrial automation systems. Buratto states that “SiC-based semiconductors can operate at higher temperatures than silicon-based devices and should be used for the next-generation of temperature sensors”.
Fu stated that “future sensors will integrate computing capabilities, enabling local data processing and intelligent decision-making to reduce latency, bandwidth use, and cloud dependency, while AI/ML will allow sensors to self-optimise based on environmental conditions”. For example, sensors could adjust their sensitivity with changes in temperature and humidity. Fu also notes that “large language models (LLMs) will improve industrial automation by enabling natural language interfaces for interacting with sensors and equipment, simplifying operations and maintenance”.
About the author
Liam Critchley is an independent technology journalist.
Related
AI-based anomaly detection can enhance industrial efficiency
