WIRELESS SENSOR NETWORKS: MOTION MONITORING

Motion monitoring forms a part of condition based maintenance. These include structural monitoring (monitoring bridges, buildings, manufacturing plant, etc.), medical maintenance and urban terrain mapping. The structures such as motor, buildings, and bridges are prone to wear and fatigue during their lifetime and have typical mode of vibrations, acoustic emissions especially during wear and tear.

WSN can continuously monitor these structural environments and report if there is any anomaly in the working of these structures. For instance, sensor networks can be used to monitor the structural changes of a bridge. The sensors are embedded beneath the road surface of a bridge and are powered by the piezoelectric crystals. When heavy vehicles move along the surface, the sensor nodes are powered.

Let us take another example. In a large semi-conductor manufacturing plants, there are thousands of routine machinery. For efficient plant operation, the machines need to operate smoothly. Any flaw in a particular machine may halt the production process. A team of electricians carry a computing device which attaches itself to the sensors in various machines and logs the data on to a central computer. The team would then analyze the data and check for signs for wear and tear. Months elapse between visits to a particular machine.

WSN offers a better approach- monitor the machinery through sensors, perform local data processing at each device and transmit the data to the operations staff. What distinguishes this operation from environment monitoring is the fact that in motion monitoring, the no of samples are extremely large-100 Hz for vibration analysis and several kilo Hz for acoustic analysis. Such a scenario poses several constraints on WSN. At this high sample rate, data processing needs to be done at faster rate demanding more power and buffer storage. Rather than transmitting the raw data, the sensor nodes can perform signal analysis and transmit only detected anomalies in the system. Reducing the cost of obtaining and processing data reduces the overall cost.

Also, the WSN for structural monitoring must operate within a time frame, i.e. they need to gather time correlated data. Nodes need to share the data or processed data at a correlated time. In structural analysis the output data from one sensor node can be used as input to the sensor nodes that are operating at critical points in the structure and thereby evaluating the overall performance of the structure as a whole. 

SENSOR NETWORKS: ENVIRONMENT MONITORING

In this post, let us take an example where sensor networks are deployed to study the environment.  When sensor networks were not around, the environmental scientists had to go deep in the field (you’d have probably seen on Discovery or National Geographic) and face the harsh conditions each time they find the need to record new observations.  But, with WSN the scientists can not only gather data frequently but also with much leisured ease.

For instance, monitoring the microclimate throughout the volume of redwood trees, helps form a sample of entire forests. Redwood trees are so large that the entire ecosystem exists within their physical envelope. Climatic factors determine the rate of photosynthesis, water and nutrient transport, and growth patterns. Substantial variations are known to exist over the volume of an individual specimen, and researchers believe that the microclimatic structure varies over regions of the forest. In addition, water transport rates and the scale of respiration may influence the microclimate around a tree, effectively creating its own weather. These factors greatly influence the habitat of the species existing in and on the trees.

Earlier, the scientists used a winch to reach the top portions of these trees and record their measurements manually. But, these results were not conclusive because of the fact that their measurements would be at a particular time of the day and using just one sample of the reading it is indeed hard to judge how the climate would be progressing throughout that particular day.

Following is the figure of a node in WSN used for environment monitoring.  

An entire weather station is of the size of a film canister. Two light sensors at the top measure the total solar radiation, especially the photosynthetic radiation bands at which chlorophyll is sensitive. There are also sensors at the bottom which measure the temperature, relative humidity and barometric pressure. These nodes are kept in the shade which protects them from rain.

The weather protected node also houses a computing device, battery, a miniature transceiver to collect data, process it and route the information through other nodes to the outside world. Using these sensor nodes is far more effective to collect data across different elevations of the tree for prolonged period of time.

Following is the the temperature profile over three days, collected from 16 nodes at four elevations in a 35-meter study tree. The WSN samples climate data every five minutes and computes an average temperature at each elevation. The measurements show that within the expected daily cycle, the top of the tree experiences much wider climatic variation than the forest floor.

The network data also illustrates the progression of temperature wave-fronts in a single day. It also gives pertinent information about the relative humidity. It has been found that redwood trees move huge volumes of water in a single day creating powerful temperature and humidity gradients that could be instrumental in understanding the growth dynamics, nutrient transport and water intake of such large structures thereby iterating the dense instrumentation.

Initially, researchers can intensively study the environment by analyzing the results. In the future, they can reprogram the WSN remotely to monitor information pertaining to deviation from regular climatic data. This would increase the lifetime of the WSN significantly.

SENSOR NETWORK APPLICATIONS

The sensor networks use intelligent sensors and pervasive networking technology to instrument the physical world which has in turn increased the density of instrumentation and also quality of information being processed.

By pervasive networking, we mean the ability for devices to (semi-) autonomously arrange themselves into local networks and exchange information through these networks. These devices are also known as ubiquitous computing devices or pervasive computing and are integrated into everyday objects and activities. In essence, pervasive computing is formally defined as devices which fit the human environment instead of forcing humans in theirs.

Using this concept, the wide range of uses for which WSN (Wireless Sensor Networks) can be put into can be broadly differentiated as:

·         Monitoring space,

·         Monitoring things, and

·         Monitoring the interactions of things with each other and the encompassing space.

Monitoring Space- This category includes Environmental and habitat monitoring, precision agriculture, indoor climate control, surveillance, treaty verification, and intelligent alarms.

Monitoring things- The second category includes structural monitoring, eco-physiology, condition-based equipment maintenance, medical diagnostics, and urban terrain mapping.

The third category forms the critical applications which involves monitoring complex interactions, including wildlife habitats, disaster management, emergency response, ubiquitous computing environments, asset tracking, healthcare, and manufacturing process flow.

WIRELESS SENSOR NETWORKS: INTRODUCTION

Before, I actually begin; I wish to recall the movie “Twister”. Hopefully, many of you would have watched this movie. If not, no worries. In this movie, a particular group of scientists fight their way into the mouth of a tornado to release sensors to gather data about the interior of a tornado. At that instant, I realized that they were actually releasing were Wireless sensor networks capable of collecting data about the tornado.

Observation is crucial for scientific progress because it forms the basis of any study. Advances in semi-conductor technology have spurred the growth of devices with large computing power and lost cost. Every year, the size and the cost of these devices are decreasing exponentially. Researchers are using this technology to gather data more efficiently and thereby creating a new dimension of computing in science.

With better fabrication techniques, it is now possible to integrate sensor, computing device and a networking device in a single system. These systems are highly proficient in interacting with the environment and collect more information which effectively increases the quality of any study.

Systems are built with miniature radios and exceptionally small mechanical devices which sense the quality of physical world. Such systems are inexpensive and are densely deployed capable of coordinating with each other. Combining such systems with system software technology has made instrumentation of the world possible with high fidelity.

To realize this venture demands more challenges. The individual systems in the Wireless Sensor Network (WSN) are inherently resource constrained:

• The computing devices operate on low memory, limited processing speed and low communication bandwidth
• The desire of WSN to be autonomous places a severe energy constraint. 

So, the design for such systems has to be in a way that these systems process information and give maximum output on an aggregate, not individually.

In most WSN settings, in order to minimize energy consumption, most of the device’s components, likely the radio are turned off. Also, the individual nodes are subjected to harsh environmental conditions. Furthermore, since the nodes are densely deployed, there will be a high degree of interaction between them and there is a greater need for developing architecture and protocols which allows their efficient operation.

Despite these challenges, the nodes must remain inexpensive and must organize themselves to program and manage the network as a whole.