Putting wireless sensor networks to work
Sensor networks are built from an infrastructure of local sensors, a communications medium, and a central, common data processing facility. A wireless sensor network builds on this concept by allowing the untethering of the sensors from the bounded medium. This allows a lot of freedom and flexibility in the placement of sensors and the ability to fine tune the monitoring capability of the network. A wireless sensor network (WSN) is the natural outgrowth of the advances made in wireless technology, miniaturization, and batteries.
This technology also is driving the proliferation of consumer grade sensors and devices that are the basis of what is popularly called the "Internet of Things" (IoT) that is capturing the public’s imagination. This article will focus on the basics of WSNs and how to use them. Readers of Control Engineering have read how wireless local area networks (WLANs) can be used to great advantage by increasing efficiency and reducing lost time; WSNs will expand this concept even further by leveraging existing wireless (and wired) infrastructures to build extensive networks of sensors that will reach into almost every area of life.
Wireless sensors typically are small, self-contained, low power units with a modest amount of processing power. The key concept to this technology is the capability of wireless communication, either with a central portal or as part of a mesh using other devices to extend range. Wireless sensors rely heavily on power saving algorithms to remain operational over the long term. Battery technology is improving; increased battery capacity, coupled with the capability of going dormant for extended periods of time, allow for expected battery life into several years of operation.
Most mobile devices already have this capability—Bluetooth and Zigbee, for example—as power saving algorithms are a required feature in all Wi-Fi certified devices, as well as in devices conforming to the IEEE 802.15.4-2015: IEEE Standard for Low-Rate Wireless Personal Area Networks (WPANs) wireless standard. The concept is simply that if there is no activity or event to report, the sensor "goes to sleep." Upon an event or a predetermined time period, the sensor wakes up, assesses the situation, reports status, and then goes back to sleep. This cycle also can be triggered by a polling algorithm that addresses each sensor in turn. Duty cycle also can be adjusted to switch the sensor on and off, effectively cutting power consumption in half. The point is that these sensors are designed from the ground up to operate as low power nodes.
A matter of size
The most significant feature of the new generation of wireless sensors is their size. Sensors are called by names that conjure up distinct images: "smart dust," "commercial off-the-shelf motes," or simply "motes." Their sizes range from nanoscale to macroscopic. The former describes biological or small passive sensors, which may or may not be embedded; the latter refers to larger sensors, such as toll collection tags, access cards, and the like. The idea is to deploy an infrastructure of small, low power, low bit rate distributed sensors with varying degrees of computing power that will form larger, high resolution, almost organic networks.
Data processing will be done by conventional means by fixed central data processing facilities being fed data from the network and performing the bulk of processing. Data processing within the network itself also is being contemplated, creating, in a sense, a "distributed processor." The eventual realization is a large distributed network communicating with and between a wide variety of sensors that will operate autonomously. This will require a common, open communication and data standard to ensure seamless interoperability.
Applying wireless sensor technologies
The list of applications for which wireless sensors can be used is quite long. A common application is security systems. Wireless access control and area monitoring are already done extensively by wireless technology. Another is the national power grid, or more familiarly, the nationwide network of weather stations. The common aspect in these already realized sensor networks is the capability to communicate over existing communication infrastructure. However, many systems are highly specialized and employ proprietary communication schemes, adding greatly to cost and complexity, hence the need for a standardized approach.
It is envisioned that future sensor networks will communicate over a large mesh architecture, with each node having the ability to forward data from other nodes, eventually landing at an aggregator where the disparate data will be processed. Sensors will be deployed in high density networks and in large quantities. They will be inter-networked using short-haul, low-power wireless links between nodes, while the existing communications infrastructure, particularly WLANs and Internet connectivity, will be used for long-haul communication.
WSNs will facilitate the instrumenting and control of homes, factories, treatment plants, ships, airplanes-the list is endless. With ubiquitous sensor networks, every facet of operation can be monitored, allowing a building, for example, to report on a structural weakness, or a treatment process to spot anomalies that would be missed by fixed traditional sensors. In an industrial process, wireless sensors could be injected into the process stream and continuously monitor thousands of points. Smart sensors that monitor several different parameters also could perform as miniature laboratories, sending data back to a SCADA system for required action. Even the human body can be instrumented using this technology, and will be extensively used to monitor such things as blood chemistry, or the body’s complex electrical activity, communicating this data to the patient’s doctor and warning of potentially harmful events. The list of applications for WSNs is endless.
Finally, a word of caution: as we have seen with the explosion of mobile devices, it is easy to become dependent on them and lose our ability to think for ourselves. WSNs are not a replacement for critical and analytical thought and action. The current cyberwar that is raging in the background should be considered thoroughly. Just as conventional networks can be compromised by a well-placed worm or Bot, so can a WSN be rendered useless, or worse, dangerous, using common intrusion methods. While WSNs offer great promise in our ability to understand and improve our respective specialties, if not our world, becoming too dependent on technology also can invite catastrophe. With ubiquity comes vulnerability.
Daniel E. Capano is a senior instrumentation project manager at Gannett Fleming Engineers and Architects in New York. He is a Control Engineering Editorial Advisory Board member. Edited by Jack Smith, content manager, CFE Media, Control Engineering, firstname.lastname@example.org
- Wireless sensors typically are small, self-contained, low-power units with a modest amount of processing power.
- The most significant feature of the new generation of wireless sensors is their size.
- Wireless access control and area monitoring are already done extensively by wireless technology.
What applications in your enterprise are good candidates for wireless sensor networks?
See related articles linked below.
Wireless research online from Control Engineering is available at www.controleng.com/ce-research.