Using wireless for closed-loop applications

When considering wireless for control applications, including it in the system design early can prevent surprises during or after installation.

By Justin Shade, Phoenix Contact January 28, 2016

Automation of industrial applications is becoming common throughout most industries. Users endeavor to maximize process efficiency and minimize costs. As technology evolves, engineers strive to find new ways to test and use automation to their advantage.

When it comes to controlling manufacturing processes, automation continues to be an enabler with a rapidly increasing influence. For example, wireless technology is a segment of automation that has matured over the period of about a decade. Although it was not considered appropriate for control at first, wireless has quickly evolved from sensor-only applications to closed-loop control.

A wireless system can become an integral part of a control system regardless of that system’s architecture: central, distributed, or closed-loop control. However, before introducing wireless into a control application, the different wireless technologies—and how they can impact control applications—must be fully understood. 

Initial considerations

Before diving into a project that involves using wireless in control applications, taking time to think about what’s ahead can prevent issues later on. The primary issues involve expectations of a wireless network, its criticality (use case), and actually choosing a wireless system—wisely. 

Expectations: When considering a wireless system for a control application, it is important to understand what the expectation of the wireless network is and to plan the system around those expectations. Wireless products are becoming faster and more reliable. However, many users expect a wireless product to function like a standard wire. There will always be a latency factor in a wireless system that a wire does not have.

This is a fundamental point to consider when designing a system. If this point is discussed during the design stage of a project, it is typically not an issue. The problem comes when latency is overlooked, which can create problems while trying to start up the system.

The wireless manufacturer can typically help guide the user on what to expect regarding latency. The system design engineer must determine if that latency is acceptable for the specific application. Typically, these latencies are not very large, but they must be a point of discussion when implementing a wireless system. 

Use case: Even with using wireless for control, there are different use cases for the wireless system. At a high level, they can be divided into two groups: critical and convenience.

With critical applications, the wireless link is being used to send critical real-time data. Typically, these data are deemed "critical" if they are time sensitive. If, for whatever reason, the wireless link is delayed or broken, the application process will be interrupted.

A convenience application is established when a wireless system is implemented to help make a control process more efficient but isn’t necessarily critical to the application itself. An example of this is networking remote programmable logic controllers (PLCs) or control devices that were historically monitored by someone physically visiting them multiple times a day to take readings or recordings. A wireless system can eliminate that need, but if the link is lost, the local control can still run. 

Choosing a wireless technology: When adding a wireless technology to an application, it is important to know that no wireless technology fits every application. While some technologies may overlap in application scenarios, they have different attributes that allow them to fit well into different applications.

When determining which radio frequency (RF) technology is right for a specific application, there are three main criteria to examine: environment, connection medium, and data throughput. When examining environment, obstructions and other RF sources in the area and distance must be considered. As a rule-of-thumb, the lower the frequency, the less the RF energy is attenuated by air. Therefore, relatively lower frequency RF can travel farther and better penetrate obstructions. The frequency used also impacts data rates. At lower frequencies, the channel sizes used to send information are smaller. This in turn lowers the data rate into the kilobit per second (kbps) ranges.

Conversely at higher frequencies, the channel sizes are larger, allowing more data to be sent-typically in the megabits per second (Mbps) range. As wireless distances increase, lowering the operating frequency may be required. This in turn lowers the data rate. This change may affect the design of the system or application.

Last is the connection medium between the system devices and wireless module. This is important because not all wireless devices have all types of connection media. Some have strictly discrete input/output (I/O) connections, while others have serial or Ethernet connections. Still others have combinations of those. Before choosing a wireless product for a specific application, it is important to verify that the product being considered has the desired connectivity.

The aforementioned criteria can help users determine if wireless technology is a possibility in their applications. After these issues have been addressed, the next step is to determine how the control system works. There are different control architectures. Based on the architecture, some wireless technologies fit better than others.

With all control architectures, the communication still must be transferred by some medium. Historically, short distance connections have used various cable combinations routed through conduit to a centralized control point. The "tech-in-the-truck" concept has been employed for longer distances where cables cannot be run. Both of these types of networking can become very time consuming and costly, depending on the installation and manpower required. With wire and cable prices continuing to rise—and companies looking to make their processes more efficient—new ways of networking systems are being investigated. A growing trend in control applications is wireless communication. With wireless communication, companies can now reduce installation cost and time when networking devices, locations, and information that they previously could not. 

Types of control architectures

Control architectures fall into three general categories: distributed, centralized, and closed loop. However, when applying them (which can involve combinations of these), some applications can become quite complex. 

Distributed: An application with a control device, such as a PLC, or control system deployed at or near the process it controls, is called distributed control. A wireless link is sometimes used to only send updates of the local control device. In this manner, it is reporting only periodically, and all the control is done by the local control system. This way, if there is some delay in response from the wireless system or a periodic break in the wireless connection, the local system can continue on its process and just send its updates when the wireless link recovers. This is a very popular use of wireless communication because if using technologies, such as 900 MHz frequency-hopping spread spectrum or a licensed ultra-high frequency band, the devices being networked together can be thousands of feet to miles away. These types of technologies are built for long-range, slower communications, typically in the kbps range. They are meant to be robust and reliable to help solve the problems of distance and obstructions. This type of wireless link eliminates the need to periodically send a person to monitor that control device. 

Centralized: Control applications are called "centralized" when there is a central control device, such as a PLC or PC that sends commands to remote devices to tell them what to do. These end devices are usually closer to the main control device, ranging from a few feet to a couple thousand feet. Having this closer proximity opens the opportunity to use wireless communication because many types of wireless technologies are available for these shorter distances. Shorter distances allow higher frequencies and higher bandwidths. Technologies, such as wireless local area network (WLAN or Wi-Fi), Bluetooth, and ZigBee, are built for short distance and allow for higher bandwidths with data rates in the Mbps range.

With these higher bandwidths, more devices can now be networked, getting rid of the need for cable and conduit and reducing installation costs. The concern in this type of application is that the use case is more critical. Because there is typically no intelligence at the end devices, they require real-time communication with the control system to send and receive its commands. If that wireless link is interrupted for any reason, it could cause the system to stop or fault. 

Closed loop: Closed-loop control is a mixture between centralized and distributed control but adds a feedback loop from a sensor to compare it to what the controller is reading. Like a centralized control system, the response of the required communication must be reliable and robust due to the number and timing of comparisons being made in the control system.

For example, consider a domestic clothes dryer. When a certain temperature is selected to dry the clothes, the dryer heats up to that preconfigured setting and holds the temperature for the period of time the dryer is on. In a standard, centralized control system, the temperature could be correct, but it is not known for sure because there is nothing reporting the actual temperature inside the dryer. In a closed-loop control system, there is a sensor in the dryer monitoring the actual temperature and sending that back to the main controller to compare and make adjustments if necessary. This allows more specific and accurate measurements within the system.

When adding wireless to a closed-loop system, the primary concerns are the importance and frequency of the information that the remote sensor must read from the main controller. In some applications, that feedback loop is critical to maintain temperatures, flows, and levels. One missed data packet could be the difference between a good and bad product (see Figure 1). In other cases, if this feedback loop isn’t acquired for seconds or minutes, it won’t affect the process at all. It depends on the application at hand.

For highly critical feedback loops, such as a centralized control system, WLAN, Bluetooth, and ZigBee are the most popular technologies due to the throughputs available (see Figure 2). Depending on how critical the individual feedback loops are, each loop may require its own wireless link to the controller, so it has a dedicated path.

Closed-loop systems can be used in long-range applications as well. Feedback loops don’t have to be in close proximity to the main controller. However, it is important to know that the time it takes for that feedback loop to get to the controller could be longer (maybe seconds instead of milliseconds). As long as this is addressed in the design stages and the criteria are acceptable, there is no reason it cannot be done.

Wireless design considerations

Historically, most, if not all, control-type applications have been wired installations. As wireless technology has grown in acceptance and has proven reliable, more control systems are being implemented using wireless (see Figure 3). Adding wireless to an application can help save wiring, maintenance, installation, and manual monitoring costs. For in-plant, short-range applications, technologies are becoming increasingly more flexible, so adding multiple dedicated wireless links isn’t as costly as it once was. Sometimes, the distance can be short, but the physical path to run the cable isn’t possible. In other cases, routing the cable might be technically possible but very expensive, meaning that a wireless system is the best option.

When considering wireless in a control application, it is important to design it into the system early so there are no surprises during or after installation. Make sure expectations and requirements make sense and the chosen product can actually function as desired. When adding wireless in a retrofit situation, running the wireless system in parallel as a test is recommended. This gives a baseline of how the system will work and allow time to determine if it is a right fit.

There are many different wireless technologies available, so when applicable, contacting the manufacturer to discuss the application, the requirements, and possible obstacles can increase the likelihood that the implementation is successful.

Justin Shade is product marketing specialist, wireless, for Phoenix Contact. He started his career at the company in 2006 in the Technical Service Department before moving to his current position in the I/O and Networking Marketing Department in 2011.

This article appears in the Applied Automation supplement for Control Engineering 
and Plant Engineering

– See other articles from the supplement below.