A smart sensor is a component, which enhance the functions for generating correct representation of the data and performance. A smart sensor enables a sensor to integrate with signal processing electronics in single compact package. It has become easier, cheaper and faster to design a sensor system and resulting systems are more reliable and provide a higher performance than traditional system. It may be dynamically programmed as per requirement. This will decrease expensive application specific sensors and general-purpose sensors. Smart sensor implemented in a distributed smart sensor system consists of a network backbone. Fiber optics and Embedded systems these components enhance the functions of smart sensor. Bluetooth technology implements wireless communication and fast and secure transmission of data with in a given range.
INTRODUCTION: -
‘Smart sensor’ you heard about them. What are they? It’s only a phrase that many engineering community always use? We would like to explain what does this phrase means. By taking a simple analogy I will introduce you to smart sensor.
In animals nerve endings are used to sense an environmental effects, such as heat light and then signals are sent to brain via nerves, where they are processed and decision is taken how to react. If a reaction is necessary, a signal is sent via another nerve and then animal responds to the environmental.
Similarly sensor is made smart by duplicating the essential elements of the above human system that consists of:
1.Embedded sensors (nerve endings)
2.Data links (nerves)
3.Programmed data processor (human brain)
4.Actuators (muscles)
A smart sensor is dynamically programmed as per user requirements. This will decrease
need for expensive application of specific sensors. Therefore smart sensors will significantly reduce the cost and time required to design, repair or modify sensor systems. This is certainly beneficial to traditional user of sensor technology such as military, industrial and automotive. Its real impact will be to make sensor technology available to a much wider range of users.
SMART SENSOR
Smart sensor is the heart of an instrument. It can be defined as device, which converts non-electrical quantities such as pressure, temperature, etc. into a useful electrical sign. Now, the most significant development in sensor technology is the inclusion of a dedicated microcomputer in the sensor design as an integral component. Such a sensor is called as “smart sensor”. A smart sensor provides a function beyond those necessary for generating the correct representation of a sensor control quantity.
The sensors are called ‘smart’ when some circuits and sensor technologies are mixed either monolithically or in the same package. SMART SENSOR = SENSING ELEMENT + SMART CIRCUIT Example:
Imagine a spectrum of intelligence: On one end of a simple thermister a semiconductor device whose resistance varies predictably with temperature and opposite end of spectrum, sensor with such powerful combination of capabilities that it eliminates the need for anything else in the system. This device calibrates itself; equalizes records and scales incoming data; compute static; reacts to the readings which are taken by altering the process. It also monitors via actuators or other controls and communicate some of the knowledge across the network of a supervisory system.
Sensor becomes smart when it does either one or all following functions:
• Performs logic function
• Performs two way communication
• Makes decision
• Smart sensors are able to compensate changes in variables to maintain accurate performance.
It means smart circuit is added with sensing element and result is ‘smart sensor’.
Calibration is also done by the smart sensor.
Features of smart sensor: -
1.Faster signal conditioning
2.Higher signal to noise ratio
3.Small physical sizes and high reliability
4.Higher performance
5.Self testing
6.Auto calibration
7.Failure detection and penetration
8.Executive command and logical function
ELECTRONIC INTEGRATION FOR SMART SENSORS: -
A general smart sensor system block diagram is a closed-loop system consisting of five blocks: Sensor, Signal Processing, AD, Digital Decision Logic, and DA.(Fig.1)
Sensor: -
It is a basic element of above diagram. It senses a physical quantity. The sensed signal can be due to change in resistance, capacitance or frequency. For example in PT-100 resistance is changes with respect to temperature. Each of these sensors requires different detecting circuits, but many blocks are common to any sensors.
Signal Processing: -
In any sensor applications, the first thing to be done is to modify the quality of the detected signal. One such enhancement is to improve the signal to noise ratio by filtering. In many sensor applications, we have preknowledge about the operating frequency range of the sensor signal. The noise is usually of wide-spectrum, by only allowing the sensor signal over its operating range using band-pass filtering will increase the signal to noise ratio. For example in biological application, the desired sensor signal is usually superimposed with the periodic heart bit signal. This periodic signal can be eliminated using band-reject filter. Continuous time filter is used for elimination of noise introduced by switching in digital filter. There is continuous time universal filter which can be used for elimination of noise and increase the signal to noise ratio. This filter is called universal because same filter becomes low pass or band reject by simply changing the external connections. A typical universal second order filter schematic is as shown in
figure. Also signal amplification us usually require to build up signal level to required input dynamic range of analog to digital converter.
Analog to Digital Converter:-
To exploit the digital decision logic to create smart sensors, an analog to digital converter is necessary. For complete integration, we need to be able to create our own AD converter. The design should be high-speed, and can be of any size or precision to accommodate a wide-range of applications. In some cases AD converter is developed based on neural network architecture that achieves a successive approximation type AD asynchronously. The maximum speed of the AD =N*T Where N is the number of bits &T is the propagation delay of the comparator. This design was successfully implemented using wide-swing high- speed comparator, and using the technology with high-res layer.
Digital Decision Logic:
Decision making to create a family of smart sensors is best delegated to the digital world, where hosts of digital devices can be used such as micro-controller, microprocessor or etc. Today’s inexpensive small microprocessors are increasingly making there way on to sensors.
Microprocessor equipped smart sensors can perform other conversions as well. Measuring devices operate in terms of secondary variable- it may be electronic, mechanical or optical. For example a thermister resistance changes with temperature. Resistance is the thermisters secondary variable. With the aid of small microprocessor a smart sensor can convert that to primary variable (that is temperature) algorithm. Similar algorithm-based means can convert data to useful engineering units (converting volts to degree Celsius)
Imagine a smart vibration sensor monitoring a piece of rotating machinery. There are hundreds of measurements about a process or a piece of machinery might be paired down to a single bit of knowledge (for example- the bearings are worn; replace them).
Digital to Analog converter: - Digital to analog converter is used to convert digital information into analog signal this is used to complete feedback path.
TECHNICAL DESCRIPTION:-
A distributed smart sensor system consists of a network backbone, on which reside many nodes. Nodes may loosely classified as
1) Sensor node
2) Control node
Sensor nodes are those, which tend to send data to network. Controller nodes are those which gather data from network. There may be more than one control node on a network. (Fig.2)
Fig.2 Smart sensor system
Prototype sensor node on a network consist of three elements (Fig.3)
1. A physical transducer
2. A network interface
3. Processor or memory core
Fig.3. Prototype smart sensor node
The transducer senses the physical quantity being measured and converts it into an electrical signal then signals are fed to an analog to digital converter and now ready to use the processor .The processor will perform some signal processing on the data and depending upon how it is programmed, may send the resulting information out to the network. Network transmissions are handled by the network interface block.
Prototype controller node consist processor or memory, a network interface and input or output devices for communicating with user. It is used to collect information from the sensor nodes, to program the sensor nodes and to provide feedback to user.
One may allow sensor nodes the ability to program other sensor nodes, resulting in highly complex system behavior. But at this point this capability is not central to the research.In above smart sensor system, placing all the sensors on a common network bus enables ‘plug and play’ facility of installation. That is no new wires have to be routed to accommodate new nodes. The processor, memory on board and the sensor nodes are essential for plug and play functionality.
In traditional system each sensor has varying gain, offsets, hysteresis, etc. which must be compensated for elsewhere in the system. A smart sensor node would store the physical attributes of the transducer and would compensate for nonidealities locally in the processor. This allows the replacement of sensors without need for recalibration.
The processor may also be used for data conversion. In many applications one does not care about the raw data, but only information derived from it. For example if one may not care about what exact temperature at a location is, but he is interested in whether it has exceeded a certain threshold or not. Instead of sending a stream of temperature readings, a smart sensor node would send just one message when temperature criteria are met. Thus only desired information is sent out to network. Here the benefit is that bandwidth required to support each node decreases, so allows more to reside on a single network. Also a processor on a smart sensor network may be considered as a distributed parallel computer.
Traditional sensor systems on the other hand, have only one central data processing resource. So smart sensor system is capable of data manipulations.
SENSING FOR SMART STRUCTURE:-
To become a smart structure, the structure must be able to sense internal conditions or states which will be the first critical function that must be fulfilled. So the ability to sense internal state should present in all smart structures. Generally there are number of different ways to sense any particular parameter or field. Some of them as follows:-
1. The sensor can be located at a point monitoring parameters of interest.
2. Extended measurement can be made if a long gauge length integrating sensor is used. For a sensor of this type, the gauge length of the sensor is distributed. The region of interest and its output represent some integrated measurement parameters.
3. Sensing can be done by using a array of point sensors. Basically it is implemented by interrogating each sensor separately.
4. The number of communication links for this discrete point
sensor array can be reduced from many to one at the expense of additional complications and sophistication through the use of multiplexing in the sensor signal processing.
5. Also sensing can be carried out in a fully distributed manner.
DETAILS OF SMART SENSOR COMPONENTS:-
Optical Fibers for smart sensors
Fiber optic sensor offers embedded sensor capability and natural connections to fiber optic data links that can be used in a wide variety of composite materials to act as the nervous system. Optical fibers for smart sensors have the following advantages
a) They are very small in diameter, only upto 125 mm. that can be embedded in many types of composite structure without changing mechanical properties.
b) Fiber optic are environmentally rugged and are able to withstand the temperature and composite structure
c) They are immune to electromagnetic interference eliminating the bulky and costly shielding.
d) Fiber optic sensors may be multiplexed so that many sensors lie along a single fiber line.
A composite panel may have embedded or attached fiber optic. Sensors that are used to monitor an environmental effect. This sensor can be multiplexed and their signals are carried by fiber optic to separate out and process the information. The data is then formatted and transmitter to a control system that may be used to convey the information to an actuator system that is directed to respond to the environmental effects.
Embedded System
Embedded systems are devices to control, monitor or assist the operation of equipment, machinery or plant Embedded signifies that they are the integral part of the systems. Thus, they are comparable to brain.
Embedded system categories:-
1) Individual microprocessor
These may be found in small devices such as temperature sensors, gas detectors, and circuit breakers.
2) Small assemblies of microprocessors with no timing function
These may be found in flow controller signal amplifiers, position sensors and value actuators.
3) Computer systems used in manufacturing or process control
This relates to causes where the computer is connected to plant or machinery in order to control it. In such systems, the computer is rather than fir direct control of individual devices within it, which almost certainly involves other kinds of embedded systems.
These systems are liable to be affected in exactly the same way as commercial data processing systems; because of course the hardware and the system software are same. So the applications software may have been developed along similar lines.
4) Microcontrollers
These are found in domestic and consumer products. These can not be programmed and rarely have a real time clock.
5) Large scale systems
These are computers with a hard disc drive connected to equipment.
6) Test hardness
These are devices intended to mimic the operation of a plant or a device. So that a human tester may determine whether the computer will effectively and safely controls the actual plant or device.
Embedded system development
Earlier engineers used to develop system design by interconnecting standard products on printed circuit boards. Since then integration levels have continued to climb and many functions that previously were built up using multiple discrete components are now integrated into a single chip.
New model for designing is called‘embedded standard products’. In this model blocks of intellectual properties are implemented as customizable ‘hard’ functions surrounded by user configurable logic. Thus, hardware and software have developed a lot.
TECHNOLOGIES FOR SMART SENSOR
BLUETOOTH TECHNOLOGY:
Designer at Erickson put up the idea of bluetooth technology is nothing but the technology in 1994.The bluetooth technology is nothing but technology which allow user to make effortless, instant wireless connections between a wide range of communication devices such as telephones, computers & its
peripherals PDA etc. So bluetooth supports both point to point and point to multipoint connection. Because of this feature Bluetooth Technologies with small RF Tranreceivers are used for implementation of smart sensor system.
Let us see example of smart pressure sensor (Fig.4). The pressure sensor is connected to the bluetooth node. This system consists of pressure sensing element, smart signal conditioning circuitry including calibration & temperature compensation &; Transducer Electronic Data Sheet (TEDS). This system architecture can be easily developed for specific sensor configurations such as thermocouples, strain gauges & other Sensor technologies. This system also includes sensor signal conditioning as well as communication functions.
The pressure sensor node collects data from multiple sensors & transmits the data via Bluetooth wireless communications in the 2.4
GHz base band to a network hub or their Internet appliances such as computers, laptop or any other hand held device. So this system is advantageous for fast & secure transmission of data within a given range (upto usually 10 meters).
Micro electro mechanical systems(MEMS)
Sensors using this technology are created from microscopic mechanical structures ranging in size from micrometer to several millimeters on silicon wafers, using IC processing techniques. The structural material, configuration and the type of coating used with the structure determine the sensor characteristics and the type of measurement that can be made.
MEMS have been extended to the use of nana-technology. In this, the physical properties change with external stimuli, as laser beam. This technology helps to manipulate atoms individually and place them exactly where one needs to produce the desired structure that is rearrangement of atom.
MEMS is a technology that combines computer with tiny mechanical devices such as sensor,
valve, gear, and actuators embedded in semiconductor chips. MEMS device contains micro-circuitry on a tiny silicon chip into which some mechanical device such as sensor has been manufactured. Such chips can be built in large quantities at low cost. Perhaps the most promising platform for smart sensors are Microelectromechanical systems(MEMS) in a MEMS device sensor itself becomes silicon, as in the case of silicon pressure sensor using a deflection diaphragm. By putting a sensor on a chip, you can place additional functionality onto the die at a little or a no additional cost. The promise of MEMS is the total integration of measurement, computation and communication all on same die.
ATTRIBUTES OF SMART SENSOR: -
Self-diagnosis
There are two types of information, one concern the normal status of the instrument and the other provides diagnosis of fault or unusual process conditions. Both are primary interest in plant maintenance and are therefore, considered together. The ability to call up this information assists in keeping an up –to- date plant
instrumentation maintenance log which an be checked by actual interrogation to ensure that the type range etc of the plant mounted equipment is in line with the plant records as well as the model number, serial number and materials, where this critical is permanent. Other details are entered or modified when the transmitter is first Installed or removed from service. Diagnostic information is concerned with actual operation of the transmitter and changes much more frequently. Some diagnostic routines prevent discrepant configuration information from being entered. Others warn of unexpected situations, such as reverse flow situations while others may report internal failures in the electronics viz., failure to write to a memory location. Many of these routines run continuously and set a flag immediately a fault is detected. The operator from the control room may request others.
Signal conditioning
The presence of memory and computing power at the transmitter permits the signal to be conditioned before onward transmission. For example, the primary sensor signal may have known non-linear relationship with the measured variable. In those cases, where the
measurement system comprises two discrete units, such as a primary sensor and a separate transmitter, which incorporates the computation unit, provision would be made to enter the parameters of the primary sensor into the computation unit of the transmitter, so that either unit could be changed. Also these mathematical manipulations of the original data to auxiliary internal measurements may be made of line pressure or temperature to permit the output to be corrected for the effect of these quantities on the sensor performance.
Multisensing capability
If we want to manage both the pressure and temperature sensing mechanisms in the same sensor then the cost saved on that account would be very significant. It is of course very lucrative to include more and more of the parameters in the same sensor. Chemical and biological sensors will be of interest in this area. The concept of multisensing potentially addresses the issue of selectivity. An array of discrete sensing elements, fabricated by various means, could sense the full spectrum of the target species. From the response of each element, the unique signature of the species could be determined by smart processing algorithms. This technique relies on
the inherent cross-sensitivity of the entire individual sensing elements. Applications would be vary right from the coffee makers to the explosion detection devices.
Capability of decision-making
Capability of decision-making is the primary objective of smart sensor. This is related to measurement function. But if the measurement can be carried out for making small decisions so that crowding of information at the next stage is reduced, then it will be useful addition. This will make energy and time savings.
Distribution of information
Information accumulated from a sensor is not just processed in a single point. It may be required to transmit to more than one point in the plant for various purposes, at right time. To accomplish this, it is better to have transmission from sensor rather than from control room because it will save much time and energy. Moreover today’s process environment is networked completely. Hence, sharing of information is more relevant and a compulsory option. Distribution of information is therefore a variable, possible solution for smartness.
Single chip realization
Saving of space is need of modern technology and smart sensors are not exception to that. Single chip realizations will make the sensors to be compatible with the digital technology and field buses and hence, will increase cost saving.
APPLICATIONS:-
Smart Infrared temperature sensor:-
In the past, if process engineers needed to change a sensor’s settings, they would have to either shut down the process line to remove the sensor or try to manually reset it in place. But today, they can remotely configure, monitor, address, and control the process using advance sensor system such as smart sensor system.
Smart infrared sensors can be used in any manufacturing process in which temperatures are crucial to high quality product. For example fig. shows a converting process with an oven, a corrugating or embossing die and a cooler.
Six infrared temperature sensors can be seen monitoring product temperatures before and after various thermal processes and before and after drying. The smart sensors are configured on a high-speed multidrop network and are individually addressable from remote supervisory computers.
Measured temperatures of all the sensor locations can be sent individually or sequentially for control purposes. Using remote addressing features, set points, alarms and signal processing information can be downloaded to each sensor. So the process control is achieved.
As smart sensors are used in this system so communication with sensors becomes very easy. There is no need to remove the sensors from process line. Smart sensors contain EPROMs, so user can reprogram them to meet their specific process requirements.
So with the help of smart infrared sensors engineers can keep the process efficient and the product quality high. As these sensors are digital processing components and communication capabilities provide a level of flexibility and safety.
Sensor Web:-
What is sensor web exactly?
With the presence of cheaper, miniature and smart sensor, abundant fast and ubiquitous computing devices, wireless and mobile communication network and autonomous and intelligent software agents.
In short the sensor web offers full dimensional and full phase sensing and monitoring of earth at all levels: global, regional and local.
The sensor web is a revolutionary concept towards achieving collaborative, coherent, consistent and consolidated sensor data collection, fusion and distribution.
Sensor web
The confluence of rapidly evolving sensors, computation, telecommunication & real time positioning technologies introduce a new instrument concept –‘Sensor Web’. A sensor web is generally composed of multiple instrument /processor platform & intelligent sensors that are interconnected by
means of communication fabrics for the purpose of collecting MSMT & processing data from various sensors (both in situated & remote) & accurate analysis are informed.
Decision Making
Talk ever. With those cheap & powerful embedded Sensors are able to think and talk. Because of advantages of semiconductor technologies, computing processors, the small sensors have become small intelligent agents. Those sensor quipped brains are able to think with the integration of computing processors, wire/wireless communication & embedded operation systems. These sensors can even with each other.
FUTURE TECHNOLOGY NEEDS: -
Much progress has been made in sensor technology, but still there are lots of changing problems, which remain to be, overcome before smart sensor become commonplace.
First &; perhaps most important is the further development of truly compatible sensor circuit fabrication processes so that the overall input can be maintained at a high level and the total cost can be made low.
The fabrication cost of a sensing element should be minimum in high performance sensor applications, the unit cost of smart sensors is not as important as the level of performance achieved by sensors, while in high volume sensor applications, both high performance & low cost are desired
Packaging of smart sensors is one of the most challenging problems facing the sensor community .In the area of micro actuators, drive mechanisms capable of producing high displacement simultaneously, are badly needed.
The increased availability of communication and networking systems (both wired and wireless) and the demand for ubiquitous communication is likely to bring about a crossing over price and functionability between sensor networking technology and communication technology.
CONCLUSION: -
Thus we can conclude that smart sensors are more technologically developed than dumb sensors. Their microscopic size and added advantage which helps to apply these sensors in different fields, ranging from medical instrument to large scale engineeringand business transudations and military weapons.
Download this Seminar Report
0 comments:
Post a Comment
Thanks for your Valuable comment