The ability to make real-time decisions based on pressure, temperature, or flow measurements while a process is running can provide significant advantages in a measurement and control system. These advantages can be expressed in different ways such as cost savings through improved resource management, or reduced reliability upon mobile workers. While measurement data can be used to dynamically control a process, it can also be displayed over a network to allow remote monitoring of the process status in real time. Rising pressures, temperatures, or vibration intensity can easily be adjusted if the appropriate personnel are aware of faulty conditions. As data is collected for process control or a SCADA system, it can also be archived for future reference when a review of process trends could provide additional improvements.

Figure 1. Wireless Technologies Comparison
While wireless sensing clearly offers many advantages, the adoption phase moves at a slow pace in many industries due to inaccurate perceptions of the technology. One of the main reasons wireless technology is yet to be fully adopted is price. Many organizations have not taken the time to explore the benefits achieved via wireless sensing in comparison to the price of implementing the technology. Similarly, a company may be content with their existing wired system and be reluctant to make the switch.

The second issue stems from concerns with the technology itself – namely, latency. This element is considered to be critical, and can vary depending on which wireless technology is chosen by an organization. However, latency is an inherent trait of all wireless technologies. Unlike a conventional analog signal, there are delays associated with the analog-to-digital conversion process, as well as the RF transmissions. The time it takes from the start of the conversion process when the raw measurement is taken, until the wireless signal from the sensor is received at the gateway or modem, is considered to be the total latency time. The latency for each wireless technology will differ. Having said this, a wireless sensor network will never be as fast or timely as a standard wired sensor network. Determinism that occurs between each layer of the wireless protocol (network, security, application, etc.) each takes a different amount of time, and this is all generally longer than its corresponding layer embedded in the wired versions of the protocols. Many wireless protocols also have to maintain very strict timing structures to make sure data packets are routed and delivered to the proper locations on time. Maintaining this timing will add to the total determinism and latency of the system.

Figure 2. Meshing Diagram
Each wireless protocol will have its advantages. For instance, the ZigBee protocol offers low latency within the millisecond range and the ability to scale a very large network size. Bluetooth will have slightly higher latencies, but significantly higher data rate capabilities, whereas WirelessHART offers channel-hopping and the ability to interface with hundreds of thousands of HART sensors and equipment that may already be located within the field. That being said, a single WirelessHART network has a maximum size of about 100 devices.

Security is another concern when considering the adoption of wireless technology. While it may not be of the utmost concern in the commercial or industrial arenas, it certainly becomes a key issue amongst the military and medical fields. In many of these applications, it would be inexcusable for data to be lost or obtained by the wrong parties. The 128-bit AES encryption is used across the majority of wireless platforms. There are several precautions that can be taken on top of this in order to further protect the wireless data, but it should be noted they do not offer advantages over standard encryption; they simply add a simpler layer of security on top.

For instance, the ZigBee protocol, like any other protocol, has a well-defined packet structure. Portions of this packet structure can be formatted in a very unique fashion by making alterations to the Application Layer of the protocol. Only those who know the exact format of these packets will be able to properly decode the data. It should be noted that the AES encryption, for example, will contribute to the overall latency by making the wireless transmissions slightly longer in duration. This will, in turn, increase the demands of battery consumption.

Benefits of Wireless Sensing

Although the most pressing issue that companies face is the question of overall reliability, what happens if a node fails? Is the data stored anywhere to be retrieved at a later time? How will the technology function in a particular environment? Without explaining every possible scenario, we can look at several specific features of wireless technology that can help eliminate many of the apparent concerns.

Figure 3. ZigBee Wireless Networks
Signal strength is one of the best indicators of reliability. If a strong signal exists, there will be no loss of communications. However, if the signal weakens, this may raise concerns regarding loss of communication links and, more importantly, loss of data. There are a few factors that can affect the strength of a wireless signal, including interference with

devices that produce a large amount of electrical or RF noise, as well as a wireless device’s ability to coexist with other wireless devices, perhaps of a different protocol such as Wi-Fi.

A common practice to ensure a customer will receive strong signal strength in their working environment is to conduct a field test. This is done by the manufacturer and consists of going into a typical working environment and positioning the wireless devices in common locations where heavy interference may occur. The devices would be streaming data packets back to a gateway device, and the signal strength of these data packets would be monitored. Purchasing a wireless sensing solution would be con-

tingent upon a successful field test. For example, imagine there are steel barriers in direct line of sight from the wireless sensor to the gateway device stored in a control room or CNC machine box. If the signal strength is strong and reliable under these conditions, then it is safe to believe that the wireless signal will be accurate under all conditions in that particular environment. In contrast, if the signal is weak and/or no signal is found, then alternative options would have to be explored.

One option would involve the use of a higher gain or directional antenna to improve the signal strength. Also offered by many wireless technologies is the ability to use a “meshing” system. Meshing is the ability to use routers or repeaters to extend the wireless signals and make communication paths more reliable. When a wireless sensor network exists with meshing capabilities, various communication paths can be created, as seen in Figure 2. In the partial mesh topology, nodes are only connected to certain other nodes, but in a full mesh topology, every node is connected to one another. This way, if a device loses communication, perhaps by loss of battery power, an alternate communication path will automatically be created, greatly reducing the chance of data loss. While Bluetooth operates primarily on a master/slave relationship, ZigBee and WirelessHART support meshing. Figure 3 shows a common configuration of a ZigBee mesh network consisting of end devices, routers, and coordinators. Coordinators and routers can communicate with any other device, but end devices can only communicate with routers and coordinators, not other end devices.

A typical WirelessHART network configuration is an example of a full mesh network topology. In addition, ZigBee has a “retry” metric that offers the ability to resend data if the original message was not properly sent and received the first time. In ZigBee, there are proper handshaking sequences that need to take place, such as acknowledgments between the wireless device and the gateway device when a data packet is sent and received. If an acknowledgment does not occur on both ends, it is assumed that the data was lost, and then is resent. This is a parameter that can be configured similar to other ZigBee parameters, such as a communication channel and transmit power, by using a piece of utility software.

Many of the features mentioned above have an effect on the battery life of the wireless sensing devices. One example of this is by turning on the 128-bit encryption that tends to consume slightly more battery life than if it were off. Similarly, the latency and/or response time of the wireless sensors will also have a large effect on the expected battery life of the device, as seen in Figure 4. Many wireless sensors on the market today have programmable transmission frequencies; in other words, it is easy to ask the device to transmit a reading once every minute or ten times per second. Considering that actual RF transmission consumes the most battery power, one can expect to see a significantly shorter battery life from a device that transmits ten times per second. For this reason, a consumer may choose a ZigBee enabled device. ZigBee is known for its ultra-low power requirements. The transmission rates are programmable, as is the actual transmit power. ZigBee devices also resort to a very low power mode when the device is not transmitting in an effort to conserve as much battery life as possible.

Signal Receiving and Control

Figure 4. Battery Life Chart
It is imperative to understand the receiving portion, also known as the gateway device. Gateway devices are used to receive the wireless signals, but more importantly, they process the data to allow interface with a company’s existing control equipment. These gateways offer numerous hardware interfaces such as serial RS-232, USB, and digital I/Os or relays, as well as software interfaces such as MODBUS-RTU or TCP/IP, OLE for Process Control (OPC), or FTP and Telnet. Common applications involve connecting a gateway device to a CNC machine’s controller through an RS-232 connection. Wireless pressure sensors can monitor clamping pressures on tooling pallets inside the machines to verify that parts are being machined properly. These measurements would be provided to the machine’s controller through MODBUS-RTU. If the clamping pressure is too low, the machine’s cycle would automatically cease, preventing the destruction of accompanying tools and/or machine parts. An error like this could result in tens of thousands of dollars to replace.

Another option would be for temperature measurements to be wirelessly transmitted to a gateway device. This data would be processed and published to a company’s network or existing historian software over an Ethernet connection having an OPC interface.

Many companies have shown great interest in having the ability to control all of their sensing needs via the

Figure 5. ZigBee Message Over IP
Internet. ZigBee, for example, is one of the first wireless standards to adopt a functional IP protocol and set of IP connected devices that can interoperate natively with other IP connected devices. ZigBee’s proprietary IP protocol, implemented in the Application Layer, is designed to easily integrate existing ZigBee devices. The standard ZigBee data packet structure is combined with a conventional IP data packet in Figure 5. The goal is to define a compact, low-traffic message format that can support embedded systems that are attached to low-bandwidth, low-power networks. Devices are allowed to communicate with one another by implementing manufacturer-specific application profiles. These profiles are a set of parameters that define communication channels and other vital settings to assure proper communication and operation between devices. Used primarily in the home automation and smart energy markets, the ZigBee/IP profiles also allow for larger sensor networks and increased network security.

Another main focus of this protocol is to meet the needs of extremely resource-constrained devices, usually running on microcontrollers. Unlike HTTP or XML, which require more resources to decode and process, the binary messages created by ZigBee are simpler to implement and are much better suited for the low-bandwidth networks.

This article was written by Aaron LaJoie, Electrical Engineer, at Electrochem Solutions, Clarence, NY. For more information, Click Here 


NASA Tech Briefs Magazine

This article first appeared in the July, 2010 issue of NASA Tech Briefs Magazine.

Read more articles from this issue here.

Read more articles from the archives here.