Sensors & Transducers Journal
(ISSN 1726- 5479)
Vol.114, March 2010, pp. I-VI
Sensors: Smart vs. Intelligent
International Frequency Sensor Association (IFSA),
Tel.: +34 696067716, fax: +34 93 4011989
According to Global Industry Analysts, Inc., the world smart sensors market is projected to reach $7.8 bn by 2015. It means that the global demand for highly integrated smart systems will increase dramatically in the years to come. Even though the economic crisis is dominating, according to ‘Smart Sensors: An International Market Report' (BizAcumen, Inc.) sensor networks and smart sensors are being used widely in industries, including automotive, medical, industrial, entertainment, security, and defence due to increased usage of process controls and sensing elements in different sectors. Use of smart sensors in counter-terrorism applications including cargo tracking, biometrics among others is also increasing. Smart systems are being readily accepted by the auto industry as automakers explore ways to save lives, deflect safety concerns and prompt costly litigation. Smart systems can sense an impending crash, its nature, the presence of occupants, their position, and determine the type of airbags to be fired and the force and speed of their deployment. Motion-tracking sensors and inertial sensors based on micro-electromechanical system (MEMS) are widely being used in medical applications. Implantable devices that are small in size and yet possess high reliability, and handheld devices for use in diagnostics and home monitoring mainly drive the demand for such sensors.
Strong growth expected for sensors based on MEMS-technologies, smart sensors, sensors with bus capabilities and embedded processing. Sensors based on MEMS and NEMS technologies and intelligent (smart) sensors are at the focus of the current sensor development. MEMS technologies allow to miniaturize sensors and, at the same time, to integrate their sensor elements with microelectronic functions in minimal space. MEMS technologies make it possible to mass-produce sensors more and more cost-effectively while improving their functionality and miniaturizing them. Thus, major sensor and electronics manufacturers expect that within a few years, there will be complete sensor integrated on one chip and having its own local intelligence for information processing.
But what does it mean a “smart sensor” ? In some languages “smart” and “intelligent” are translated from English by the same word that normally means “intelligent”. But in English there is a difference. It is a reason of existing of two widely used smart and intelligent sensor definitions. The first one is more related to technological or so-called “smart” aspects; the second one is more related to functional or so-called “intelligent” aspects. The first definition was formulated by the following way: smart sensor is a combination of a sensor, an analog interface circuit, an analog to digital converter (ADC) and a bus interface in one housing [1, 2]. It was the first smart sensor definition and is used widely till now. A smart sensor can be made as integrated (if all elements of smart sensors are integrated into one chip) or hybrid sensor. The second definition was formulated by the following manner: intelligent sensor is the sensor that has one or several intelligent functions, and is used often in various references. There was also an attempt to merge both “technological” and “intelligent” aspects of smart sensors in a single definition : smart sensor is one chip, without external components, including the sensing, interfacing, signal processing and intelligence (self-testing, self-identification or self-adaptation) functions. Some interesting smart and intelligent sensors definitions in its developments were described in .
What does it make a sensor to be “intelligent” ? Very often it means a presence of microcontroller or microprocessor. However, it is a necessary but not enough condition. Sometimes a microcontroller is used only for calculation according to predetermined equations. Speaking about intelligent sensors various intelligent functions mean a self-testing, self-validation, self-checking, self-diagnosis, self-adaptation, self-identification, self-calibration, self-compensation, self-etc …
In common case, a smart sensor can not have any intelligent function, as well as an intelligent sensor can not be a smart. In other turn, an intelligent sensor system can be based on any sensors, not only on smart or intelligent.
2. Survey Results
In 2009 a mini survey “What does it mean “smart sensor”?” was set up by IFSA at Sensors Web Portal. We get very good feedback from 227 participants from both: academia and industry. The survey results are shown in Fig.1.
Speaking about other suggestions made by survey’s participants, first of all it should be a sensor with extended network capabilities.
The survey´s results justify that modern sensors and sensor systems include more and more intelligent functions and are based on modern microelectronics and MEMS technologies. The interest to smart and intelligent sensors and systems is continuously growing up that is demonstrated by the current survey 2010: “What is your topic of interest in Sensors & Transducers journal ?”, where during the first month we get reply from more than 400 participants, and “Smart sensors and systems” topic was selected among most popular topics of interest .
3. Future Perspectives
Modern microelectronics and MEMS technologies and its rapid developments bring new possibilities for integration of all smart sensors elements, in other turn, novel methods of measurements and advanced algorithms introduce more intelligence into smart, intelligent sensors and systems. The aim of this editorial article is not to formulate a precision, modern definition of smart (intelligent) sensor (it is a call for standardization), but rather to distinguish further perspectives and modern trends in smart (intelligent) sensors development.
From technological “smart” point of view an IC-compatible 3D micro-structuring technology is being developed . An integrated smart sensor should contain all elements including wireless communication and power management. Without doubt, innovative technology concept such as smart systems integration will play a crucial role in this respect. Smart systems, defined as intelligent, often miniaturized, technical subsystem with their own and independent functionality evolving from microsystems technology. Smart systems are able to sense and diagnose complex situations. They are “predictive”, they have the capability to decide and help to decide as well as to interact with the environment .
The greatest progress in innovation, however, will happen where MEMS technologies overlap with smart technologies. For intelligent sensor systems this means a trend from miniaturized sensor to the smart and miniaturized sensor systems, which can integrate processing functions in a minimal space. The main goal of smart sensor development is to improve the reliability and durability of these sensors and make them more easily adaptable to new functions and conditions during the operating phase. In addition to self-diagnosis, self-calibration and self-validation capabilities, smart sensors must have the self-adaptation function. Self-adaptive intelligent sensors can adapt themselves automatically to the measurement task according to current measurement conditions and algorithms. The main goal in bus sensor development is to make them more easily adaptable to communication networks (including wireless networks) and control systems. According to the ‘plug and play’ principle, installation in complex measuring and automation systems can be made much easier; the intelligence of the system as a whole can be decentralized, and the sensors can be used in a multiple way by the various control systems.
Most effort in recent years has focused upon the advancement of the sensing devices and associated hardware, increasing the sensitivity and reactivity of these systems. These capabilities however cannot be fully used in any cost-sensitive application, as this traditional solution to the measurement challenge is several orders of magnitude more expensive than a smart-sensor should cost or can be afforded in most applications (high performance ADCs cost several dollars). However there is a need for improved sensors with increased accuracy, reliability and acquisition speed.
The ability of new intelligent sensors systems to process information is not enough: efficient coupling this ability with decision making based on data processing, on advances in data structures and upon the knowledge of each application environment in order to learn and adapt will be required.
There are a lot of new technologies, suitable for smart sensors creation. But still there is a problem how to joint them and use in a frame of intelligent sensor systems on chip.
Advanced packaging technologies, such as System-in-Package (SiP) technology (Fig.2), is gaining importance in the smart sensors systems integration. With its significantly higher flexibility compared to System-on-Chip (SoC), the SiP can combine multiple technologies and reuse intellectual property from numerous sources, allowing designers to overcome integration difficulties without compromising on individual chip technologies. In other words, SiP-based smart sensors systems offer numerous advantages – more functions in less space, reduced design cycle times and time-to-market.
Fig. 2. System-in-Package (SiP)
(Source: Connolly C., Miniature electronic modules for advanced health care
Sensor Review, Vol.29, No.2, 2009, pp. 98-103).
Fully integration of all blocks and functions into a SiP is expensive, and only a revolution approach will make the cost per integrated smart, intelligent sensor reasonable. How to achieve this important goal ? Nevertheless the Strategic Research Agenda of the European Technology Platform on Smart System Integration  describes well “what” should be done, but “how” and in “which way” are till not determined, especially in smart sensors and systems area. Even the usage of more advanced technological processes, for example, 45 nm CMOS for SoC design, it will be slow evolution but no revolution in design because of it is related only on a further scaling, and as usually, bring initially many problems at production: low yield rate, high price, etc., what we have in a few last decades. Where is a solution ? It is in a revolution approach to smart and intelligent sensors design.
The innovation, revolution approach for intelligent sensors and sensors systems creation must:
Most modern sensors and transducers have analog output with voltage, current or charge, proportional to a sensing parameter. In order to convert an analog signal to digital signal different kind of analog-to-digital converter (ADC) are used. However, there is an alternative approach, when a frequency-time parameter of electrical signal can be used as an informative parameter on sensor’s output. In this case frequency, period, duty-cycle, PWM, phase-shift or pulse number in such so-called quasi-digital sensors is proportional to measurand and a frequency-to-digital converter should be used to obtain a digital output. Quasi-digital sensors are rather interesting from a technological and fabrication compatibility point of view, the simplifications of the signal conditioning circuitry and measurand-to-digital converter, as well as metrology performances and the hardware for realization. The last one essentially influences on the chip area.
Moving from the traditional analog signal domain to the frequency-time signal domain lets achieve many benefits due to properties of frequency as the informative output parameter for sensors, namely: a high noise immunity; high output signal power; wide dynamic range; high accuracy of references; simplicity of signal switching, interfacing, coding and integration; multiparametricity. Using the frequency as the output signal for sensors is an extremely useful alternative to the conventional analog voltage output signal and it can be easily accomplished with relatively few components. By eliminating the need for an ADC, frequency output sensor schemes reduce the cost of sensor systems. No output standardization is necessary as in the case of analog signal domain. Many type of sensing elements and read-out circuitry can be merged by this way on a single chip or in SoC or SiP. In additional such approaches will give a great opportunity to create new self-adaptive smart, intelligent sensors and systems due to the use of novel, advanced methods for frequency-to-digital conversion.
The importance of microsystems is continuously growing because of the combination of two trends: the progress in silicon sensor technology and the introduction of new techniques for interface circuits. The cost of measuring systems has greatly reduced due to the batch fabrication of both the sensors and the interface circuits. Interface circuits are the most critical part of the signal processing chain. This means that the overall performance of the system strongly depends on the quality of the parameter-to-digital converter. In spite of the availability of 24-bit resolution, high speed and low-cost ADCs the standard design approach relies on more expensive mixed IC design. This raises cost and yield issues. An alternative approach is to accurately convert the period or (frequency), PWM or duty-cycle of an incoming square or rectangular wave signal to digital with timing resolution better than the typical ADC resolution. As usually, such approach is related to standard CMOS digital design rather than expensive mixed or analog design. In addition, main components such as microcontroller core, precision voltage-to-frequency converters, Universal Sensors and Transducers Interface  and Universal Frequency-to-Digital Converter  are exist or can be easy realized as a standard library cells in various CAD tools such as Cadence, Mentor Graphic, Cyprus and other. Sensing elements and miniaturized resonators can be realized with the help of existing MEMS technologies.
In addition such innovation approach will open up new horizons also in nano-electromechanical systems e.g. sensors/resonators/arrays; high-performance nanomechanical oscillators fabricated by bottom-up integration of silicon nanowires for applications such as sensitive mass detection and radio-frequency signal processing.
Without doubt, innovative, revolution technology concept will play a crucial role in the coming future. Smart sensors will be smarter and intelligent sensors will have more artificial intelligence, and finally, we will get real smart, intelligent sensors.
.Huijsing J.H., Riedijk F.R., van der Horn G., Developments in Integrated Smart Sensors, Sensors and Actuators, A: Physical, Vol.43,
.Huijsing J.H., Smart
Sensor Systems: Why ? Where ? How ?, in Smart Sensor Systems, ed. by
Gerard C.M. Meijer, John Wiley
.Kirianaki N.V., Yurish
S.Y., Shpak N.O., Deynega V.P., Data Acquisition and Signal Processing
for Smart Sensors, John Wiley
.Taymanov R., Sapozhnikova K., Problems of Terminology in the Field of Measuring Instruments with Elements of Artificial Intelligence,
Sensors & Transducers, Vol. 102, Issue 3, March 2009, pp.51-61, available online at:
.Survey 2009: What is your topic of interest in Sensors & Transducers journal ?
.Strategic Research Agenda of the European Technology Platform on Smart System Integration, Version 2, March 2009.
.Yurish S. Y., Smart Universal Sensor and Transducer Interface, in Proceedings of the SENSOR+TEST Conference 2007, Nurnberg,
Germany, 22-24 May 2007, Vol.1, pp. 307-312.
S. Y., Novel Universal Frequency-to-Digital Converters and Sensors
Interfacing Integrated Circuits, Sensor Electronics
Click <here> or title of paper to download the full pages article (667 Kb)
1999 - 2016 Copyright ©, International Frequency Sensor Association (IFSA). All Rights Reserved.
Home - News - Links - Archives - Tools - Standardization - Patents - Marketplace - Projects - Wish List - Subscribe - Search - e-Shop - Membership
Members Area -Sensors Portal -Training Courses - S&T Digest - For advertisers - Bookstore - Forums - Polls - Sensor Jobs - Site Map