A new generation of high-performance industrial sensors support secondary pointing accuracy and precision geo-location while providing the necessary size and cost efficiency, and are beginning to work to drive the development of 'removable IoT'.
With the increasing popularity of high-quality sensors, combined with reliable connectivity and data analysis, new industrial efficiencies have been created that also make these smart nodes more autonomous and actionable. In many cases, sensor nodes Precise motion capture and location tracking become the main core of successful applications.
For example, smart farms can use more ground and air traffic with automated geolocation of sensor content and analytics to guide ground operations more efficiently, and smart operating rooms bring typical navigation to the operating table , A sensor-guided, precision-guided robot ensures accurate guidance under all conditions In many areas, motion-based sensors add value to the value of a mobile application.
Consumer inertial sensor applications are already widely used in smartphones, but they also cause users to generally find their accuracy less than satisfactory; thus, in the concept of driving the Internet of Moving Things (IoMT) Has been ineffective.However, a new generation of high-performance industrial sensors can support the sub-degree of accuracy (sub-gegree) and precise location of geolocation, while providing the necessary size and cost-effective, and is beginning to promote IoMT development of.
Intelligent Sensing Driver in Industrial Systems
The most valuable advance in industrial machinery and manufacturing is to focus on tangible, system-level benefits that often result in design and implementation challenges and, on the other hand, translate into new solutions and business models.3 There are currently three This type of system-level driving force has been designed to pursue resource efficiency, criticality and security, and applications focused on these enhancements span different industries across the air / land / ocean, indoor / Outdoor, short-term / long-term and human / mechanical, but in any case, these applications rely on common attributes; that is, accuracy, reliability, security and intelligent processing and analysis, as listed in Table 1.
Table 1: Valuable System Attributes translate into challenging design requirements in IoMT applications
With multiple types of sensors at the heart of the target application design task, however, developers must carefully consider the quality and durability of the sensor in a wide range of changes in response to the complexity of the design, although some industries may choose based on convenience Sensors (using, for example, existing sensor kits in handsets), but other industries will define sensor kits from scratch and select them based on precision and intelligently integrate them for complete and reliable coverage of all anticipated systems status.
Intelligent sensing
These smart and accessible systems have revolutionized the already mature industries with sensor-rich contextual architecture, transforming agriculture into smart agriculture, upgrading infrastructure to smart infrastructures, and urban transition to smart cities. Environment-related situational information and deploy more sensors to databases that require cross-platform and cross-time integration (for example, how the infrastructure over the past year, such as crop yields or traffic conditions and patterns, are analyzed over the cloud) Management and communications (as opposed to mere sensor-to-sensor) also bring new levels of complexity, as shown in Figure 1.
Solutions that reliably extract information from equipment and the environment are the key metrics for the ultimate utility and growth of these innovative companies, with accuracy driving efficiencies that translate into the necessary economies of scale and are central to safe and reliable operations. While adding simple features is not difficult for the most basic sensors, this minimum added value is replaced by more sophisticated solutions for targeted IoMT applications (yes / no, up / down or on / off, etc.) ) And the additional influence on the choice of sensors are actually inadequate.
'Action' is crucial
In most cases, the IoT is in motion, precision pointing is still crucial even when it is not dynamic (such as a stationary industrial security camera), or it is valuable to realize that unwanted actions (such as tampering) occur. For UAVs that use optical loads to capture crop images, better results can be provided faster, provided accurate horizons can be maintained under harsh flight conditions, and if the optical data can be precisely geo-referenced Mapping, but also for historical comparison of data and trends.
GPS vehicles are increasingly dependent on GPS navigation systems, whether on the ground, in the air or at sea, however, GPS is also at greater risk of threats, both deliberately and naturally (buildings, trees, tunnels, etc.) If accurate sensors are selected, reliable dead reckons with more sensors during outage operations can be performed Table 2 lists examples of events that add action (M) to IoMT, showing the correlation of actions to utility Sex.
If there are specific opportunities and methods that capture the natural inertia of a device or person, then the knowledge of the state of the retrieved system can be increased in importance and can be properly integrated into the existing situational information as shown in Table 3.
Reliable and secure IoMT node
The effectiveness and value of IoMT node outputs depend most on the quality of the core sensor and its ability to fetch the application context at high fidelity.Secondly, for continuous sensor calibration / enhancement and ideal sensor-to- For example, which sensor is the most reliable at any particular point in time.) The application-level processing is layered into the solution and is optimized for the details of the environment, including appropriate constraints Although these nodes are autonomous, they work together in some situations, such as assembling a group of unmanned vehicles on the ground or in the air, etc. In these cases, secure communication links are deployed and reliable transmission is enhanced As well as the unique identity protected, as shown in Figure 2.
Sensor located in autonomous core
Like the human body, autonomous IoMT nodes rely on multiple sensed inputs to realize the awareness needed for independent action and to optimize for random or even cluttered events, eventually improving over time. As noted in 4, the transition from basic measurement to control or autonomy has to be more precise at both the sensor consolidation level and the embedded intelligence. Because these nodes have a high level of interconnection and autonomous learning capabilities, Towards the human body and the direction of the polymerization machine.
No infrastructure required
GPS is omnipresent, unless the satellite signal is blocked or out of service. If it can be accessed smoothly, it can achieve extremely precise wireless ranging technology. If it can be interfered, the magnetic field reading will always exist. Inertia can also work independently. Obviously, inertial MEMS sensors have their own drawbacks (drift), but these issues are manageable and a new generation of industrial inertial measurement units (IMUs) offers unprecedented stability in a compact and cost-effective package.
Inertial MEMS components use standard semiconductor processes, precision packaging, and integration solutions to directly sense, measure and interpret their behavior, often in the form of linear acceleration (g) or angular rotation (° / sec or velocity) As shown in Figure 3. Because almost all ideal applications have so-called multiple degrees of freedom (in fact, the action can take place in any and all axes, while the device is relatively unrestricted in its actions) The g and velocity measurements must be taken separately for the x, y, and z axes or, in some cases, the roll, pitch, and yaw axes. Combining these, sometimes referred to as 6-DOF inertia Measurement unit.
Although economic factors naturally drive MEMS designers to capture multiple sense types (g, rates) on each axis (x, y, z) with minimal die space, to meet even more challenging industrial challenges In fact, some MEMS architectures attempt to measure these six modes in a single MEMS mass.It is important to check the validity of such schemes for high-performance sensing The first thing to understand is that even though MEMS elements must be used to capture motion data, it is equally important that the same elements rule out the possibility of being interpreted as other types of erroneous actions. For example, For angular rates, the effect of acceleration or gravity on angular rate measurements should also be negligible. Simple MEMS components that attempt to make a variety of measurements in a compact structure are naturally (and in design) subject to other disturbing errors Source, and therefore can not distinguish between what is needed and what is not needed, which will ultimately translate into noise and errors in navigation or pointing applications.
The IoMT, which offers promised, valuable resource efficiency, security enhancements, or critical accuracy, when needed, requires more precision than simple sensors that are ubiquitous in today's mobile devices. Performance-specific designs translate into separate Designing for each sensing mode and each sensing axis must be done in a convergence and consolidation direction Finally, it is important to understand that performance-oriented design does not need to be done at the expense of cost-effective design For the price.
Features or performance
Some applications may simply mean the substantial value of added features (in the device's attitude / direction switching pattern) that can easily be captured with simple MEMS components.Industrial or specialty devices may be easier to measure because With multiple directional accuracy and sub-level differences, or an order of magnitude higher accuracy of position recognition, and can operate in high vibration environment. Low-level and high-performance sensors is not very different between the performance Smaller, in fact, the difference is so great that it must be carefully considered when selecting components.
The end application will determine the desired level of accuracy, and the quality of the sensor selected will determine whether the target will be met.Table 5 Comparing the two solutions shows that the importance of sensor selection is not limited to the design process, but it also affects Device precision. Low-precision sensors may be suitable if they are used only in limited circumstances and have fault-tolerant applications; in other words, they are suitable for safety-independent, life-less applications or Relative precision does not need too high is sufficient.Although most of the consumer-level sensor with low noise, and can be fully operational under good conditions, but not suitable for dynamic action (including vibration) Because low performance inertial measurement units can not separate the parts they need from simple linear acceleration or tilt measurements.In order to operate in an industrial environment with better than once accuracy, The focus of sensor selection is on designs that can reject drifts from vibration or temperature effects. High-precision sensors like this It is expected to hold a wider range of application state, and may experience longer periods of time.
Designers of precision instruments are usually most interested in using inertial measurement units because their output is corrected for g and velocity rather than travel angle or distance because this system-level information is highly application-specific and therefore Became the focus of activity for system designers, not inertial sensor designers, where the challenge was to identify, for example, the pointing accuracy in inertial sensor spec sheets.
The specifications for mid-range industrial components are compared to those of typical consumer sensors found in cell phones in Table 6. It is worth mentioning that higher-level industrial components are also available, with specifications that are more The order of magnitude of an order of magnitude lower order consumer components does not include parameters such as linear acceleration effects, vibration correction, angle random walk, and other parameters that may actually be the source of maximum error in industrial applications.
The industrial sensor sample is designed for use in scenarios that anticipate relatively fast or extreme motion (2000 ° / s, 40g), and in this situation if optimal signal identification is to be achieved, a wide range of bandwidth sensors Output is also important. The minimum drift (operational stability) of the offset during operation is the desire to reduce reliance on larger complementary sensors to correct for performance. In some cases, however, the time required to correct back-end system filtering , It is also extremely important to minimize (repeatability) the start-up drift, and low-noise accelerometers are used with gyros to help resolve and correct any gravity-related drift.
Gyro sensors have in fact been designed to directly eliminate any effects of gravity events such as vibrations, impacts, accelerations, gravitational forces on component offsets, provide a substantial advantage in the form of a linear g, Temperature drift and calibration were corrected, and without calibration correction, even a typical multi-axis MEMS device integrated into a single-wafer structure may not be aligned and can be a major contributor to the error budget.
Although noise has ceased to be a major factor in discriminating sensor types in recent years, parameters such as linear g-effects and uncalibrated parameters have become sources of increased noise in any application, either through the chip design or It is the most cost-effective part-specific calibration to improve, rather than a simple or relatively static motion decision.
Sensor fusion can solve the problem of poor sensor quality?
To put it simply, No. Sensor Fusion is a filtering function and algorithm that combines and manages sensors in relation to the environment, motion dynamics, and application state. It can provide deterministic corrections like temperature compensation and can be based on system status Knowledge, which manages coordination from one sensor to another, however, does not correct native defects in the sensor.
The most crucial task in sensor fusion design is to develop in-depth knowledge of the state of the application to drive other design processes, followed by the selection of the appropriate sensors for a particular application, which requires detailed analysis to understand them Weights (correlations) at different phases of the overall mission In the example of pedestrian mapping, the solution is largely limited by existing devices, such as embedded sensors in smartphones, not due to performance design. For this reason, it becomes quite dependent on GPS and other existing sensors such as embedded inertia or magnetics, which can only make a small contribution to the task of determining valid location information, which naturally works outdoors , GPS can not be used in challenging urban environments or indoors, and the quality of other available sensors is poor, leaving a big gap or, in other words, the uncertainty of location information quality Although advanced filters and algorithms are often used to combine these sensors, there is no need for additional Sensor or a higher quality sensor but want to substantially fill the uncertainty gap, the software can do it with only minimal, ultimately reducing the credibility of the return location, as shown in Figure 4 .
In stark contrast, the industrial dead reckoning solution is designed for system-defined performance with component selection based on specific accuracy requirements and it is clear that better-quality inertial sensors have the primary Role qualifications, other sensors must be used with caution in order to reduce the uncertainty gap. Algorithms are more conceptually focused on optimal weighting between sensors, coordinated regulation and cross-correlation, and environmental awareness and immediate motion dynamics Rather than just extrapolating / estimating the position between reliable sensor readings.
In either case, accuracy can be enhanced with improved quality sensors, and although sensor filtering and algorithms are an important part of the solution, they alone can not eliminate the gap caused by limited sensor coverage.
Featuring a sensor performance nearly identical to that used by the previous generation to guide the missile, the new industrial sensor leverages the sensor architecture originally envisioned for automotive reliability and precision, as well as an economically viable and expandable process that These new generation industrial sensors are completely unique in terms of price / performance and performance dimensions, as shown in Figure 5.
Precise motion sensing is no longer isolated from niche applications and can only be invested in other costly tracking solutions. With industrial-grade precision in mini IMU sizes, IoT designers can now integrate high-quality motion sensing And drive IoMT's progress by combining embedded situational awareness to multiply the value it provides.