Wireless Sensor Networks (WSN) are here and if you're not familiar with them, you need to be. Market researcher, Onworld, estimates that the WSN market will reach $8 billion globally by 2010 - generating $5 billion from industrial applications. This high growth is driven by the ability of a WSN to lower the total cost of equipment ownership and maintenance, improve productivity and performance, and reduce labor costs across many industries.
According to an On World study on Wireless Sensor Networks, "only 5 percent of the most mission critical equipment is monitored today. The rest is monitored manually on a schedule, or not at all." Is it disconcerting that 95% of equipment is monitored manually or not at all? Yes, but given the expense of running cables to existing equipment and the cost of manpower for installation, its understandable that the return on investment (ROI) is only achievable for the most critical of assets. However, many manufacturers are beginning to realize the benefits of wireless sensors, which include:
continuous monitoring to optimize process operations;
faster ROI based on lower installation and labor costs.
Many companies have limited the deployment of wireless sensors due to the difficulties of integrating with proprietary interfaces and disparate standards. Octave Technology has built a Wireless Sensor Framework called OCTAVEX which is designed to assist in the deployment and management of wireless sensor networks.
How Industrial Sensors are Used
Industrial sensors come in a plethora of form factors and a very wide variety of monitoring capabilities such as temperature, humidity, vibration, pressure, flow, just to name a few. In industrial environments, sensors are typically used to monitor the operation of a piece of equipment, an asset, or environmental conditions. Sensors are often used in locations that are inaccessible or hazardous to workers; such as inside a fuel tank or pipeline, atop a turbine engine, or numerous other places. Beyond the obvious safety benefits, companies are realizing significant financial and operational efficiencies derived by using wireless sensors, especially when compared to traditional monitoring methods.
The Five Methods of Monitoring Industrial Equipment and Assets
In most industrial operations, there are five methods of monitoring equipment and other assets. The method chosen for monitoring any asset is directly related to the cost, importance, and location of the equipment.
The most basic monitoring involves manual sensing and manual communication. A maintenance operator travels to a piece of equipment, inspects it, manually records the results, and travels back to the control center. Ideally, the person performing the inspection is very knowledgeable of the equipment. They use their own senses (look, listen, smell, touch, and in rare instances, taste), to know if the equipment is performing within specification or not. This method of monitoring is typically used in a scenario where:
Equipment / asset cost:
Low
Equipment / asset importance:
Low
Equipment / asset location:
Varies, but the further the distance, the greater the labor cost and typically, the more infrequent the inspection.
Monitoring Frequency:
Rare: Infrequent: usually scheduled and depends on location
Initial investment cost:
None: Labor expense already in place
On-going Labor cost:
High
Accuracy:
Low - Medium: Varies based on the experience of the inspector
Problem response time:
Slow
In the second method of monitoring, one or more sensors are either built-in, or added to the equipment. A maintenance operator travels to a piece of equipment, checks the sensor(s), manually records the results, and travels back to the control center. This method of monitoring is used in the following scenario:
Equipment / asset cost:
Low - Medium
Equipment / asset importance:
Low - Medium
Equipment / asset location:
Varies, but the further the distance, the greater the labor cost and typically, the more infrequent the inspection.
Monitoring Frequency:
Rare: Infrequent: usually scheduled and depends on location
Initial investment cost:
Low: Cost of sensor and installation if not included with equipment / asset
On-going Labor cost:
High
Accuracy:
High
Problem response time:
Slow
Considerations for Manual Sensor Communication
These first two methods of monitoring use manual communication. Typically, technicians make ?rounds? through an industrial facility. Like doctors, they go from one piece of equipment to the next and checking the health of equipment and assets. The more frequent the inspection, the higher the labor cost.
In facilities where there are thousands of pieces of equipment, it's not uncommon to only inspect each asset once a month or even once a quarter. So what happens in the time between inspections? If something fails, the technician would not know until the next inspection and would have limited information to diagnose the problem for prevention in the future.
Infrequent inspections and slow problem response time is not acceptable for important or operation critical equipment. With the equipment advances and technicians becoming more "technology reliant" rather than a "hands on" approach, the need for a more automated system arises. Over time, the manual communication must be replaced by wired and wireless sensor communication.
In the third method of monitoring, one or more sensors are still used, only the operator no longer travels to the equipment to take sensor readings. Monitoring occurs on a much more frequent or continuous basis. This type of sensing is common for operational critical equipment. This method of monitoring is used for:
Equipment / asset cost:
Medium - High
Equipment / asset importance:
Medium - High
Equipment / asset location:
Varies, but the further the distance, the greater the cost to install and maintain the connection.
Monitoring Frequency:
Frequent: typically set based on equipment importance and type of monitor
Initial investment cost:
High: Cost of sensor and wiring installation
On-going Labor cost:
Low
Accuracy:
High
Problem response time:
Rapid : based on frequency of monitoring and alerting capability
Considerations for Wired Sensor Communication
An important consideration for wired communication is the installation cost, which is typically calculated as follows:
Cost of the wire (measured in $ per linear foot or meter) + Labor time for installation (often measured in linear foot or hourly) + Misc. costs (changes to facility, securing wires, ensuring compliance with safety standards)
According to the US Department of Energy, "installation of wiring can represent 20% to 80% of the cost of a sensor point". Depending on the location of the sensor this can be $200 to $400 per sensor, plus wiring cost, which can be $200 per foot depending on type of cable. Additionally the wiring can be expensive to service and maintain. These costs are one of the primary motivators driving the adoption of wireless sensors. More equipment can be monitored and more efficient maintenance procedures can be put in place.
In the fourth method of monitoring, one or more sensors are still used, only the wire has been replaced with a Radio Frequency (RF) transmitter. The transmitter uses a constant power source, typically an electrical outlet, but solar powered sensors are becoming more common. In areas where there are challenges to wiring, this scenario is used. This method of monitoring is used for:
Equipment / asset cost:
Medium - High
Equipment / asset importance:
Medium - High
Equipment / asset location:
Varies
Monitoring Frequency:
Frequent: typically set based on equipment importance and type of monitor
Initial investment cost:
Low - Medium: Cost of sensor and wireless communication installation
On-going Labor cost:
Low
Accuracy:
High
Problem response time:
Rapid : based on frequency of monitoring and alerting capability
In the fifth method of monitoring, one or more sensors are still used, only the RF transmitter is part of a sensor device. These self-contained sensor devices use batteries and the battery life needs to be a consideration. Battery life depends on how much power is required for sensing, how often the sensor device is transmitting, and how much power is necessary for the RF signal to reach the destination. In outdoor environments, there can be sensors with rechargeable batteries connected to solar panels. This method of monitoring is used for:
Equipment / asset cost:
Medium - High
Equipment / asset importance:
Medium - High
Equipment / asset location:
Varies
Monitoring Frequency:
Frequent: typically set based on equipment importance and type of monitor
Initial investment cost:
Low - Medium: Cost of sensor and wireless communication installation
On-going Labor cost:
Low
Accuracy:
High
Problem response time:
Rapid : based on frequency of monitoring and alerting capability
Considerations for Wireless Sensor Communication
Wireless communication can not only be less expensive than wired communication, it can be up and running much more quickly. Depending on the facility, it may take weeks or even months before wiring can be installed, inspected, and approved for operation. A wireless sensor can be deployed in minutes.
A couple of facility-dependent considerations for wireless sensors include:
Is there interference that may prevent the wireless sensor from working- A Radio Frequency (RF) site survey should be performed prior to making a significant investment in any wireless technology.
What is the distance between the sensor and the central monitoring location? There are different methods of communications which will be discussed later section, but typically, the greater the distance, the more power or number of repeater nodes is required.
One of the technological breakthroughs in wireless technology is "mesh networking" which allows each sensor device power up for a fraction of a second when it has a result worthy of transmitting, then relay the data to the nearest neighbor sensor. Instead of every sensor transmitting its information to a remote base station, the sensor data moves point by point until it reaches a central monitoring location where it can be recorded and analyzed. Hundreds of such sensors in a facility can provide a high number of pathways to communicate data to and from the central location. Logic built-in to each node corrects for the failure of any sensor to transmit its data by having its neighbors cover for interrupted pathways. If the communication between sensors fails, the other sensors not only report it, but can select an alternate path to the central location. See animation right.
Standards Used By Wireless Sensor Networks
One of the biggest challenges facing companies deploying wireless sensors is the disparate standards, protocols and methods of communication and data formats. Wireless network standards organizations such as ZigBee, WirelessHART, and the ISA SP100 are all working to make wireless standards and products more appropriate for industrial applications. As we all know, standards bodies are typically slow to move and it is unlikely in the long term that any one company will utilize only a single standard for all applications.
Don't misunderstand; standards are very important for the adoption of wireless sensor technology. Creation of standards typically leads to an eventual explosion of end products built on that standard which is a necessity for any emerging market. However, the confusion around competing standards and which is the best for your application can lead to inaction and market roadblocks. Interoperability across any and all standards will help drive the adoption of wireless sensors.
The other issue that exists is that there is currently no data format standard. While the methods of data packet transfer are defined in these standards, the method and format of data sent out of the network is not. Take a ZigBee network for example, the transmission rate, power consumption, and embedded stack configuration are all included in the standard, how the data is actually delivered into an application is left up to the developing company. This can add confusion for a company who is deploying ZigBee products from two vendors as the data structure may not be the same.
In the below table, we've constructed a short list of some of the standards being used in wireless sensor networking. This list will only get longer and that means more confusion for the end user. Having a platform that can create interoperability between all these devices to create an "agnostic" environment, if you will, allows end users to pick the "best of breed" wireless sensors regardless of the standard used.
Wireless Standards for Industrial Sensors
Description
802.11
A family of specifications developed by IEEE for local-area networks. Typically high bandwidth and high speed data rates and larger data packets.
802.15
A family of specifications developed by IEEE for personal-area networks. Typically low power, low rates, and small data packet size.
Bluetooth
A short range wireless standard operating in the unlicensed 2.4 GHz spectrum
ISA SP100
An open industrial wireless standard trying to support multiple protocols in a single standard. Includes 2.4ghz and 802.14.5 radios.
WiFi
Wireless Fidelity - technologies based on the 802.11 standard.
WiMax
World Interoperability for Microwave Access ? a standard for wireless broadband over long distances based on IEEE 802.16
WirelessHART
An open wireless communication standard from the HART Communications Foundation designed for process measurement and control applications. Based on the 802.15 standard and frequency hopping spread spectrum technology
ZigBee
A low data rate, two way standard for home automation and data networks. Uses very low power consumption to create mesh networks using on 802.14 radios.
Zwave
Low power, low bandwidth communications standard designed for interoperability between systems and devices. Geared toward the residential and light commercial devices.
Interfacing with Wireless Industrial Sensors
Benefits Of Working With a Wireless Sensor Framework
Collects, processes, and analyzes sensor data from any number of commercial wireless sensor devices.
Integrates with any enterprise business system, industrial communication protocol, or custom back end application.
Can be utilized by customers to significantly reduce the cost of new wireless sensor application development, eliminating the risk of competing standards, obsolete products, and application specific integration.
Supports the ability to deploy cost-saving applications that monitor equipment health, predict critical failures, and ensure vital environmental conditions, quickly and economically.
The disparity between communication and data is why a Wireless Sensor Framework is critical. The software acts as a translation layer that converts the wireless sensor data stream from disparate networks into a standard format (such as XML) in order for applications to utilize the sensor data. Otherwise, a custom interface must be built each time a new sensor device using a different data format is added.
There are numerous applications that interface with a single kind of sensor, but not multiple kinds of sensors. For example, there are applications that can monitor the level of fluid in a tank but can't combine the level data with the information from the water quality sensor. Even if both sensors communicate using the ZigBee protocol, they may have different data formats.
If you already have a software application for monitoring data from wired sensors and you're considering adding wireless sensor, we suggest you do some homework first. Adding wireless sensors to a software application designed for wired sensors can be very complicated. The application needs to understand the communication methods listed above. You may want to consider a wireless sensor framework can also handle wired sensors; one that is designed to capture wireless data and send it to the existing application and act like a wired sensor.
According to Dr. Pedro Marr'n at the University of Stuttgart, "Most applications out there at the moment are very application specific. There is a distinct need for a middleware layer to cope with the diversity of software layers."
It is often critical that manufacturers implement a software platform that can interface with a variety of sensor types as well as support both standard and proprietary communication and data formats.
A well architected platform can become the foundation on which your WSN solution is built, so it is only logical to implement one sooner, rather than later, as you begin to:
Deploy new wireless sensors;
Replace existing wired sensors with wireless;
Create new business processes, practices and capabilities for Condition Based Maintenance.
We mentioned earlier that wireless sensors can be set up more quickly than wired sensors. However, to benefit from rapid deployment, you still need a platform that allows the sensor devices to be quickly integrated with other existing systems.
Benefits for Interfacing With Sensor Devices
The OCTAVEX Framework is a software product that acquires, manages and analyzes information from industrial sensor devices.
Maintenance and operations personnel need to be able to monitor sensor data and be alerted if something occurs outside the normal operating parameters. Therefore, all sensor data should be captured, recorded, analyzed, and archived.
Sensor data needs to be analyzed in as close to real time as possible. Alert triggers need to setup based on the conditions not just from one sensor, but multiple sensors. Alert notification needs to occur via the appropriate means (i.e. text message or SMS, e-mail, etc.).
Although alerting is very important, alerts are reactive to a problem. Where companies have started obtaining real value is by preventing failures through a more proactive approach to using sensors:
1) Response-based maintenance: react when there is a problem. 2) Preventative maintenance: based on a schedule, maintenance personnel check on equipment or asset regardless of if it needs it or not. 3) Condition based maintenance: based on certain criteria, maintenance is performed. This approach provides the lowest cost of ownership for equipment, but you must be monitoring the appropriate sensor data.
A system should be open for integration with existing business systems so that it may interface with a work ticket ordering system such as a Computerized Maintenance Management Software (CMMS) or Enterprise Asset Management (EAM)or Plant Asset Management (PAM) software package. This is critical for condition based monitoring. Take this simple example: if a motor's bearings are vibrating too much, or the equipment's operating temperature gets to high, then the Intelligent Sensor Framework should automatically generate a work ticket and an alert to the maintenance team. This eliminates having to walk out and check on regular basis yet proactively prevents a potential failure. Condition based maintenance provides a lower the total cost of equipment ownership, improved productivity and performance for operations, and reduces labor costs.
Finally, consider that the more sensors you have, the more data you have to manage and platform scalability is essential. As your network of sensors increases, the ability to handle an increasing amount of data and connect to more systems becomes invaluable. Using a platform that is architected out of the box to expand and be flexible will not only improve processes today, but save time and money tomorrow.
Next Steps
There are a lot of considerations when evaluating industrial sensors and we hope that this article has provided enough useful information to get started. Throughout this article, there are breakout boxes that list some of the benefits of the OCTAVEX Wireless Sensor Framework. Octave Technology has successfully implemented a number of wireless sensor infrastructures. If you are a manufacturer, an integrator, or application service provider, the team at Octave Technology would like to talk with you. I strongly encourage everyone to take a look at their unique offering by downloading a free trial copy or scheduling an on-line demonstration by contacting Chris Parker or Jeff Thorn.