Author: Mark Scantlebury

The New Intel® Atom™ Processors for the Internet of Things

Posted on by Mark Scantlebury

In October 2016, Intel introduced its first 14 nm chips designed specifically for IoT and embedded applications. These processors—collectively known as the Intel®Atom™ processor E3900 series—offer a long list of new features for industrial, video, and other apps (Figure 1). Intel also offers an automotive-specific processor called the Intel® Atom™ processor A3900 series.

Figure 1. The Intel® Atom™ processor E3900 series delivers excellent performance and new features for IoT edge devices.

In this blog, we look at how the new features of this processor family—which includes Intel® Atom™, Intel ® Pentium®, and Intel® Celeron® processors—address the demands of IoT and embedded design. We also examine three early products from members of the Intel® Internet of Things Solutions Alliance using this new platform.

Meeting Edge Needs for Tomorrow’s IoT

With its 14 nm fabrication and embedded/IoT-specific design, the Intel Atom processor E3900 series offers a number of important advancements:

Better performance and I/O. The Intel Atom processor E3900 series delivers 1.7 times more computing power than the previous generation. Available in dual- and quad-core designs, the new processors run at up to 2.0 GHz (even higher with Intel Celeron and Intel Pentium processor SKUs).

Memory speeds and bandwidth are also improved, with support for up to LPDDR4 2400. Six USB 3.0 ports and four PCI Express* (PCIe*) ports (six lanes) expand high-speed connectivity and reduce the need for external hubs.

Upgraded graphics. A new graphics engine improves 3-D graphics performance by 2.9 times compared to the previous generation. The processors also feature enhanced image processing with low-light color processing and multi-frame technology.

Intel® Time Coordinated Computing Technology (Intel® TCC Technology). For applications requiring determinism, this new feature synchronizes peripherals and networks of connected devices. It also resolves latency issues in applications, such as robotics manufacturing, by enabling one microsecond (1 μs) timing accuracy across the network.

Improved security. A new Intel® Trusted Execution Engine (Intel® TXE) provides enhanced data and operations protection, keeping data away from hackers even if the OS is compromised. Secure boot is strengthened with features like Intel® Boot Guard 2.0, and new cryptographic instructions like Intel® SHA-NI Extensions are among the many security upgrades.

Enhanced reliability. Dual-channel ECC memory is now available when using DDR3L, helping protect against single-bit memory errors. A junction temperature range of -40°C to 110°C expands industrial use cases, and specific SKUs are qualified for automotive applications.

Fast, Strong Processing and Sensor Integration at the Edge

Three new products from Alliance members that use the Intel Atom processor E3900 series provide excellent examples of their utility and versatility in enabling more intelligent edge and fog devices.

Axiomtek ICO100 DIN-Rail Fanless Industrial System

This DIN-rail fanless embedded system uses the Intel® Atom™ processor x5-E3930 and a DDR3L SO-DIMM slot holding up to 8 GB RAM to offer a powerful solution for smart factory automation and smart energy (see Figure 2).

Figure 2. The Axiomtek ICO100 employs the Intel® Atom™ processor x5-E3930 to offer a powerful solution for smart factory automation and smart energy.

Measuring only 31 x 100 x 125 mm, the ICO100 provides a compact solution for an industrial IoT gateway. It includes 2x RS-232/422/485 ports, 2x USB 2.0 ports, 1x Gigabit Ethernet, 1x VGA, and 1x DIO. Furthermore, the unit includes 2x PCIe Mini Card slots, one for an mSATA storage card and another for 3G/GPRS/Wi-Fi connections.

The unit’s rugged construction includes an extended temperature range of -20°C to 70°C and anti-vibration up to 2G to ensure reliable operation in harsh environments. Its wide range 12V-24V DC power input includes a lockable terminal block-type connector for industrial automation application. Overvoltage and reverse protection lowers the risk of crucial data loss.

Supermicro SuperServer 5029AP-TN2 Mini-Tower

Targeted for communications, print imaging, retail, and industrial applications, this mini-tower uses an Intel® Atom™ processor E3940 to pack a lot of power and I/O in a compact design (see Figure 3). The unit features up to 8GM unbuffered non-ECC DDR3, 4x 3.5″ hot-swap drive bays, and 2x 2.5″ fixed drive bays.

Figure 3. The Super Micro SuperServer 5029AP-TN2 uses an Intel® Atom™ processor E3940 to pack serious power and I/O in a compact design.

USB ports include 4x USB 3.0 (rear), 1x USB 2.0 (Type A), plus 2x USB 2.0 via headers. A dual GbE LAN Intel i210-AT controller provides network connection options up to 1000BASE-T. PCIe options include 1x PCI-E 3.0 x16 slot, M.2 PCIe 3.0 x4 (M-key 2242/80), and 1x Mini-PCIe with mSATA.

The mini-tower offers a VGA, DP (Display Port), HDMI, and eDP display port. For extra security, there’s a TPM 1.2 aboard.

IBASE IB811 Single Board Computer (SBC)

To meet a range of power and performance needs, the IB811 SBC includes a selection of the new processors (see Figure 4). No bigger than a 3.5-inch disk, you can order it with the Intel® Atom™ processor x7-E3950, x5-E3930, Intel® Pentium processor N4200, or Intel® Celeron® processor N3350. The SBC also includes two memory sockets supporting up to 8GB of DDR3L-1866/1600 SO-DIMM modules.

Figure 4. The IBASE IB811 is a power single board computer (SBC) no bigger than a 3.5-inch disk.

The IB811 supports a wide-range operating temperature (-40°C to 85°C). A 9V~36V wide-voltage input makes it suitable for rugged industrial and in-vehicle applications with varying voltage input requirements. To comply with EuP/ErP standards, IB811 supports the iSMART green technology for power failure detection, power on/off scheduling, and low-temperature monitoring.

The IB811 supports three simultaneous displays up to 4K. Interfaces include 1.4b HDMI, DPI, and eDP/24-bit dual channel LVDS.

At the rear edge I/O are 1x COM, 1x DisplayPort, 1x HDMI, 2x GbE, and 4x USB 3.0. The board also supports 2x USB 2.0 ports via pin headers and a total of four serial ports. Other I/O includes a full-size Mini PCIe slot and an M.2 (B-key) connector.

Just the Beginning

These three examples are just a small sampling of the solutions available with the new processor. Discover more systems and boards using these processors in the Alliance’s Solutions Directory.

Bring Smart Building Tech to Smaller Facilities

Posted on by Mark Scantlebury

Smart building technology has proven its ability to reduce expenses and improve maintenance in large facilities, but these benefits have been difficult to attain in smaller buildings. The complexity and cost of building automation has relegated this technology to the largest and most sophisticated facilities and organizations. These limitations have restricted the market available to system integrators, solution providers, and value added resellers (VARs).

The Intel® Building Management Platform (Intel® BMP) aims to change this situation with an application-ready architecture that makes smart building tech simpler and less expensive. Intel BMP combines readily available off-the-shelf components and software with an extended reference design, cloud connectivity, and strong security. As illustrated in Figure 1, the platform can manage virtually all aspects of building automation.

The Intel® Building Management Platform is a comprehensive reference architecture.
Figure 1. The Intel® Building Management Platform is a comprehensive reference architecture.

The Intel BMP architecture connects disparate building equipment and devices that use a variety of protocols and sends their data to cloud-based services and applications. This connectivity enables flexible building management applications, such as dashboards for control and monitoring or data analytics for bigger-picture business intelligence (BI).

By providing the foundational hardware and software, Intel BMP enables VARs and integrators to avoid wasting time on rudimentary design chores. Instead, they can focus on higher-margin customization, consultation, and their add-on offerings and services. VARS and integrators can also help their customers move from concept to implementation more quickly and economically.

Solid IoT Foundation

The Intel BMP is built on Intel® IoT Gateway Technology, which provides a bridge between sensors and the cloud. The first gateway certified for use with Intel BMP is the Advantech UTX-3115, a palm-sized gateway equipped with a dual core Intel® Atom™ processor E3826 or E3815 (Figure 2).

The Advantech UTX-3115 is a palm-sized gateway.
Figure 2. The Advantech UTX-3115 is a palm-sized gateway.

The UTX-3115’s compact, fan-less design enables it to be placed almost anywhere. It provides connections for Ethernet, serial RS-232 or RS-422/485, USB, SATA, and TPM. A half-size Mini PCIe expansion slot for additions such as a Wi-Fi module and a full-size PCIe slot for mSATA round out the hardware.

The UTX-3115 also includes software common to all Intel IoT Gateway Technology designs. This includes the Wind River Intelligent Device Platform XT* 3.1, a robust Linux-based software stack for gateways and other IoT devices. On top of this, McAfee Embedded Control* security technologies tightly integrate the critical, hardware-based security of Intel® processors with operating system and application software security. This helps protect against cyber threats and keeps building systems safe from outside influence.

Smart Building Software

What makes Intel BMP truly useful for smart buildings is the automation software from Candi. Candi PowerTools gives cloud-based smart building applications and services secure and easy access to data and things in commercial buildings. The software provides discovery tools to locate existing sensors and a drag-and-drop interface that allows installers to quickly find and provision sensors and collect their data.

Candi’s software includes interfaces for hundreds of devices and data types over a wide range of protocols, including BACnet, Modbus, ZigBee*, Wi-Fi*, Ethernet, Z-Wave, serial, powerline, and many manufacturer-specific interfaces. There are also CANDI-compatible third-party adapters that can connect many legacy devices.

In the cloud, Candi securely connects smart buildings to leading apps and services that provide business intelligence, analytics, dashboards, reporting, remote control, and alerts to building owners and operators. The cloud portion of Intel BMP also provides the ability to remotely provision, manage, and maintain the software with over-the-air updates, such as the most recent protocol drivers, critical security patches, and new features. The update capability reduces administration headaches and helps keep building systems safe from electronic intrusion.

Smart Buildings Don’t Have to be Big

In summary, Intel BMP is a platform that gives VARs, integrators, and solution providers an enormous head start in developing smart building applications. For the vast number of small- to medium-sized companies looking to realize the benefits of smart buildings at a price they can afford, Intel BMP provides the perfect platform on which to build their solution.

Turn Data Overload into Smart Analysis

Posted on by Mark Scantlebury

Smart cities put the “big” in big data. Data needs to be gathered across huge geographies, with a vast variety of data types and analysis requirements. For example, cities may need to collect data including temperature, humidity, pressure, chemicals (pollution), movement (e.g., of cars and pedestrians), sound, and ambient light, just to name a few.

The challenges with smart city applications are many. First, data must be filtered at the source to avoid overloading networks with irrelevant data. In many cases, local analytics are required to make sense of the data. Then the data must be transported upstream across various cloud connections and through multiple APIs to a control center’s display or a decision-maker’s dashboard. In some cases, automated responses are required.

An IoT Foundation for Smart Cities

The Intel® IoT Platform was designed to tackle these challenges. It provides an end-to-end reference architecture that defines the key components of an IoT solution and how they work together. In doing so, the platform provides a foundation for securely and seamlessly connecting cities and enabling intelligent analytics (Figure 1).

Figure 1. The Intel® IoT Platform defines key components of a smart city architecture.

Most important, the Intel IoT Platform provides a common architecture that multiple IoT solution vendors can build on and integrate with. For example, SWIM.IT developed a stream-based IoT framework that integrates with the Intel IoT Platform. The SWIM framework translates data from millions of devices into stateful streams and meta-services that can be consumed in-motion by applications (Figure 2).

Figure 2. Swim.It is a secure, microservices-native, end-to-end fabric.

SWIM securely manages the end-to-end “chain of custody” of IoT data flows to ensure data integrity and regulatory compliance, making it easier to integrate IoT data into larger IT systems. Use cases include public transportation fleet management, traffic management, environmental quality monitoring, and ride sharing.

Stream Processing Makes Analytics Manageable

With millions of sensors transmitting billions of bytes of unstructured data, smart cities present unique challenges in terms of separating urgent information from mundane noise. Identifying important data in a timely fashion can have major consequences, sometimes even the difference between life or death.

To analyze high volumes of data at high speed, SAS Event Stream Processing takes a unique approach (Figure 3). Instead of running queries against stored data, it stores the data management and analytical routines, and streams massive amounts of data through these queries. This effectively allows the technology to filter, normalize, and aggregate the data and detect anomalies and patterns in realtime.

Figure 3. SAS Event Stream Processing lets users analyze high-velocity big data before it’s stored.

SAS Event Stream Processing can be integrated into other SAS applications, such as SAS Asset Performance Analytics, SAS Decision Manager, and SAS High-Performance Risk. These connections enable in-depth analysis that can trigger the appropriate action at any point across the city.

Rolling Out a Smarter IoT

These are just a few examples of the ways developers can analyze data from millions of sensors across a smart city. For more examples of sensor data analysis solutions—many designed for specific use cases—check out the Solutions Directory.

Smart Vision Secures Cities and Services

Posted on by Mark Scantlebury

Editor’s Note: insight.tech stands in support of ending acts of racism, inequity, and social injustice. We do not tolerate our sponsors’ products being used to violate human rights, including but not limited to the abuse of visualization technologies by governments. Products, technologies, and solutions are featured on insight.tech under the presumption of responsible and ethical use of artificial intelligence and computer vision tools, technologies, and methods.

 

Vision systems can fill many roles in the modern city. They can help optimize traffic and pedestrian flow, secure infrastructure and facilities, and even provide first responder “man down” alerts.

Cities began using surveillance cameras decades ago, but the emergence of intelligent vision systems is quite recent. Rapid advances in video analytics, the falling cost of high-resolution cameras, and the ubiquity of communications infrastructure are all contributing to dramatic changes in vision system capabilities. And the availability of high-performance, low-cost video processing and storage systems are making advanced systems practical for a wide range of applications.

Situational Awareness

A good example of an intelligent vision system is the LiveCast Sentinel from IoT Smart Systems, Inc. (Figure 1). The Sentinel software allows a city to track first responders both visually and with telemetry to get full situational awareness. Plus, the application enables two-way communication with field personnel via monitoring and retransmission of video and audio streams, corresponding IoT sensory telemetry, and streaming documents and pictures.

Figure 1. Livecast Sentinel allows a city to track first responders both visually and with telemetry.

Sentinel also offers bandwidth-efficient streaming optimized for UAVs and drones. This feature is invaluable for first responders trying to assess an emergency situation. For example, the streaming UAV video could be used to give an entire fire department a shared bird’s-eye view of a disaster recovery effort.

Video Analytics

Cities are also gaining new insights into quality-of-life issues through video analytics. Video analytics has progressed far beyond simple motion detection and image capture. Today, pattern recognition and deep-learning algorithms can identify objects and safety hazards, detect behaviors like loitering, and compensate for noise such as rain and fog (Figure 2).

Figure 2. Video analytics filters out irrelevant information and identifies objects and actions of interest.

These advanced algorithms require equally advanced processors like the sixth generation Intel® Core™ processors. These processors incorporate powerful graphics engines that can run vision algorithms using the OpenGL and DirectX APIs, as well as video engines for real-time encoding and decoding of IP video streams.

For example, the Intel® Core™ i7-6700K processor starts with four cores running at up to 4.2 GHz, with 8 Mb of cache. It then adds the Intel® HD Graphics 530 engine running at up to 1.15 GHz with 64 Gb of memory.

The sixth generation Intel Core processors’ capabilities are integral to the design of the Sabot SFF1 video analytics system from EmbedTek (Figure 3). The hybrid system accepts both analog and IP cameras so it is backward compatible with installed infrastructure, while providing an upgrade path to future technologies.

Figure 3. The Sabot SFF1 from EmbedTek uses sixth generation Intel® Core™ processors.

The chassis measures 9 x 13 x 3.5 inches. The motherboard and chassis were sized to accommodate an optional video capture card, while a custom small form factor optimizes airflow in higher-temperature environments.

Balancing local and cloud video storage

For municipalities or the companies serving them, deciding where and how to store the video so it can be accessed readily for analysis and reference provides a unique opportunity to innovate. Defaulting to a closed-loop, local-storage solution seems simpler, faster, and more secure. But a cloud-based solution offers more flexibility and scalability, as well as access to higher-level IP video analytics.

For designers and developers who may want the best of both worlds, Infortrend developed the EonStor GS3000 family, with unlimited cloud storage and high-performance local storage (Figure 4). This allows a municipality to easily expand a storage area network (SAN) or network attach storage (NAS) application into cloud services.

Figure 4. The EonStor GS family of storage servers from Infortrend Technologies rely upon the Intel® Xeon® processor D-1500 family.

This requires high-end processing, communications, and protocol support, for which the EonStor GS3000 family uses the Intel® Xeon® processor D-1500 family, with either two or four cores.

In total, the EonStor GS family supports 4 x 1-GbE and 4 x 10-GbE ports, with a maximum RAM of up to 256 GB. Infortrend’s “smart algorithms” allow data to be optimally allocated between the system and the cloud, which can be any private or public service, including Amazon, Microsoft* Azure*, and Google.

Infortrend lets users take full advantage of the cloud with options like Cloud Tiering, Cloud Cache, and Cloud Backup. Given the importance of security, it’s worth noting that the EonStor GS family provides AES 256-bit encryption for data-in-flight and data-at-rest, as well as self-encrypting drives (SED) compatibility, ensuring data is always protected from malicious threats. Furthermore, with integrated SSL, links between server and client are also encrypted.

In case of disk crashes, the system features integrated backup functions such as Intelligent Drive Recovery (IDR), snapshot, local replication, remote replication, and file-level synchronization.

These are just a few examples of the elements that developers can use to implement a smart city IP video security and surveillance system. For more examples of key IoT video elements—many designed for specific use cases—check out the Solutions Directory.

LoRa and LTE Cat. 1 Extend Wireless Sensing Range

Posted on by Mark Scantlebury

As more sensor nodes connect over wireless networks, there is a growing need for alternatives to interfaces like Wi-Fi*, Bluetooth*, and ZigBee*. Many Internet of Things (IoT) applications require lower cost and power consumption, longer range, and an improvement in the number of devices per router or aggregation point.

LoRa and LTE Cat. 1 are emerging as two leading alternatives. While these are proving successful as interfaces, there are still questions about interoperability, security, and manageability—and these are being addressed through gateways based on the Intel® IoT Platform.

LoRa connects many devices over long ranges

LoRa (short for Long Range) is a low-power specification designed for battery-operated devices (Figure 1). A typical interface can handle about 65,500 end devices with a range of up to 50 km and data rates up to 50 Kbit/s. Combined with localization capability, it is easy to see why LoRa is so attractive for long-range, low-power IoT sensing applications.

Figure 1. The LoRaWAN MAC sits on top of the LoRa PHY layer.

On top of the LoRA physical layer (PHY) sits the LoRaWAN media access control (MAC), which controls both the PHY layer and access to the backhaul network. LoRaWAN uses an adaptive data rate (ADR) mechanism and star-of-stars topology to ensure scalability as the number of nodes increases.

A LoRa gateway may need to aggregate data from other networks, including Bluetooth and Wi-Fi, and perform data conditioning and other number-crunching. Thus, a LoRa gateway may need high performance in addition to the aforementioned security, scalability, and manageability.

A good example is the SGWMC-X86LR-12132 Gateway from EXPEMB (Figure 2). It is designed for scalability and to support multiple interfaces and software services. In addition to LoRa, this gateway supports a 1-Gbit Ethernet link, Wi-Fi, 3G/4G, and Bluetooth, simultaneously. This gateway's support of multiple wireless interfaces is critical as not all radio protocols support IP natively, so the gateway functions as both an aggregator and IP translator.

Figure 2. The Embedded Experts gateway supports multiple interfaces.

To enable this rich functionality, the gateway is built on the Intel® Atom™ processor E3800 product family. With up to 4 speedy cores, the Intel® Atom™ processor has plenty of performance for aggregation and translation. This performance also supports the gateway's software services, including remote firmware updates and multiple layers of security, such as TLS and IPSec.

While LoRA is the newest of the long-range IoT connectivity specifications, cellular network providers are working hard to lower power and cost to provide IoT connections over licensed bands. This is important because LoRa uses unlicensed bands. In theory, using licensed bands reduces interference and makes for more reliable connections, with higher data rates.

The LE910 series of modules from Telit support one such interface: LTE Cat. 1 (Figure 3). The module supports the category's full 10-Mbit/s downlink and 5-Mbit/s uplink speeds and is optimized for both Verizon and AT&T.

Figure 3. The Telit LE910 LTE Category 1 module supports rates of 10 Mbit/s for the downlink at 5 Mbit/s in the uplink.

The module comes with IP support, as well as UDP/IP stacks and HTTP, SMTP, FTP, and SSL. It also supports a host of services such as module management that make IoT deployments under mobile networks more effective. Other features include multi-constellation (GPS + GLONASS) positioning, over-the-air firmware updates, and MIMO and receiver diversity support.

The module can be used in gateways like Quanmax UbiQ-100 Series (Figure 4). In addition to the built-in features like an Intel® Atom™ processor, HDMI, USB 3.0, USB 2.0, and COM ports, Quanmax offers integration services to incorporate custom features like LTE modules.

Figure 4. The Quanmax UbiQ-100 Series can incorporate any RF module.

The Quanmax gateway is also notable for its software options. Preloaded options include security management; remote monitoring management; wireless monitoring; data collection & translation; and more.

Wireless options for billions of nodes

As the IoT grows to billions of nodes—including countless wireless sensors—it is becoming more important than ever to make sure the right networks are applied to each application. LoRa and LTE Cat.1 are just two of the leading options for wireless sensors. To see more ideas for wireless connectivity, check out the Solutions Directory listings for wireless access and IoT gateways.

Monitoring and Securing the Smart Grid

Posted on by Mark Scantlebury

Phasor measurement unit (PMUs) play a critical role in the smart grid, helping utilities dynamically assess and manage the performance of independently operated grid systems. By keeping different parts of the grid in sync, PMUs ensure that consumers get reliable delivery of electricity, at the lowest overall cost to the consumer and the service provider.

PMUs measure the magnitude and phase angle of alternating current (AC) using a common time source, such as the global positioning system (GPS). These time-tagged measurements, or phasors, give utilities a synchronized view of performance across the grid (Figure 1). When multiple phasors are taken at the same time, they are collectively called synchrophasors.

Figure 1. A phasor measurement unit (PMU) gathers and processes the data on voltage magnitude and phase angle from powerline sensors.

To gather and analyze this data effectively, utilities need to have data acquisition, processing, and visualization, all working seamlessly across platforms and systems. To meet these demands, National Instruments (NI), Hewlett Enterprise (HPE), and OSIsoft used the Intel® IoT Platform as the foundation for a “best of breed” PMU.

This smart, connected PMU builds on the interoperability, connectivity, security, scalability, and manageability features of the Intel IoT Platform. It comes with open code that lets grid owners get deeper and faster insight into problems, so they can optimize transmission and distribution (T&D).

Configurable, high-performance data acquisition

The heart of the PMU is the customizable NI cRIO-9038 CompactRIO Platform. This platform can be programmed to have many personalities, such as power quality assurance, smart switch, or recloser. In this instance, it is set up as a PMU that can be located at any point in the grid, including on top of power poles or at substations.

A fundamental feature of the CompactRIO Platform is its combination of a processor for control and analysis, and an FPGA for hardware-based data processing. Specifically, the cRIO-09038 uses an Intel® Atom™ processor E3825 with the performance to run complex algorithms and control routines, and a Xilinx Kintex-7 160T FPGA with the configurability to keep up with changing standards and protocols.

Using the cRIO platform, processor and I/O modules can be exchanged as needed, giving designers an upgrade path as data gathering and processing requirements increase. In this case, the cRIO-09038 comes equipped with gigabit Ethernet interfaces, USB support, serial ports, and an operating temperature range of -40?C to 70?C to handle extreme environments. The whole system is programmed using NI’s high-level LabVIEW graphical system design software.

Moving in from the edge, the PMU data is passed to centralized servers powered by HPE Proliant DL380 Gen9 servers (Figure 2). The latest versions of these servers feature Intel® Xeon® processor E5-2600 v4 with 2500-MHz DDR4 memory to accelerate time to insight. These servers function as phasor data concentrators (PDCs) and are rack mountable so they can scale to handle as many PMUs as needed.

Figure 2. Hewlett Packard Enterprise Proliant DL380 accelerates time to insight.

The final stage in the synchrophasor solution is data analytics and visualization. This is the role of OSIsoft PI System, a highly secure, NERC CIP-qualified system that integrates and correlates the data with traditional operational electric grid information (Figure 3). It complies with C36.118, the IEEE standard for the measurement of synchronized phasors of power-system current and voltages. In effect, the PI System’s visualization and analytics help utilities fully understand what’s going on with their grid network, while its early-warning system helps operators react to problems quickly.

Figure 3. OSIsoft Pi System provides utility engineers and operators a variety of ways to analyze and optimize their grids. (Source: Intel)

A clear vision for the smart grid

As the smart grid continues to expand, designers are being asked to help utilities face challenges such as the distribution of renewable energy, changing load requirements, more regulations, lower cost, scalability, reliable communications, security, and fast and accurate analysis for improved operator response times.

These requirements call for smart PMUs like the one built by NI, HPE, and OSI. Their synchrophasor is end-to-end, high-capacity, validated and tested, remotely managed, and cost-effective via standards-based components. This improves service and reduces overall cost.

And of course this is only one way you can start building a smarter grid. To see more design solutions, take a look at the Solutions Directory.

Fog Computing and the Evolution of Industry 4.0

Posted on by Mark Scantlebury

As the Internet of Things (IoT) moves from buzzword to widespread deployment, it has become clear that bandwidth, storage, latency, security, and other issues are serious limitations for many systems. This realization has resulted in the emergence of fog computing, a more distributed approach to data processing, analysis, and storage that delivers analytics when and where it’s needed.

Recognizing the need for this more distributed approach, Intel, ARM, Cisco, Dell, Microsoft, and the Princeton University Edge Laboratory formed the OpenFog Consortium in November 2015. The group’s goal is to define the fog computing architecture and ensure interoperability.

Already the group is finding common ground around the system architecture specification (SAS). This specification incorporates two aspects of the Intel® IoT Platform (Figure 1):

  • Connecting “dumb” endpoints that lack integrated intelligence, security, and Internet connectivity
  • Connecting smart devices and providing real-time, closed-loop control of data shared between the devices and the cloud

Figure 1. The Intel® IoT Platform is architected to enable a robust end-to-end IoT ecosystem. (Source: Intel.)

The Intel IoT Platform is an end-to-end reference architecture and portfolio of products from Intel and its ecosystem designed to work with third-party solutions. Together they provide a foundation for connecting devices, delivering trusted data to the cloud, and enabling analytics.

Timely analytics has been a key challenge for IoT deployments. For industrial control and manufacturing applications, latency requirements are often on the order of milliseconds—and missing this window can result in catastrophic failure.

The cloud isn’t well suited to real-time analytics in these scenarios. That’s why fog computing localizes data analysis where it can be most effective, transferring only metadata, exceptions, or extreme anomalies upstream. In doing so, latencies are reduced, but also less data is transferred back and forth to the cloud. This frees up bandwidth, helps lower overall power consumption, and enhances security.

How IoT Gateways Fill Out the Cloud

In many cases, the key to local analytics is the IoT gateway. Deploying a gateway with strong processing capabilities not only makes it possible to get various “things” connected, it also enables much of the analytics to happen where these things are located.

One good example is the Nexcom* CPS 200* (Figure 2), based on Intel® IoT Gateway Technology. Designed for ease of deployment and excellent security, the CPS 200 facilitates data acquisition and exchange among industrial control systems, and between factories and the cloud, paving the way for Industry 4.0.

Figure 2. The Nexcom CPS 200 is a powerful gateway.

A key feature of the CPS 200 is its quad-core Intel® Celeron® processor. This powerful processor lets the gateway go beyond basic data acquisition to perform robust analysis on the plant floor.

Another good example is the Dedicated Computing Edge7000 IoT Gateway (Figure 3). Like the Nexcom gateway, the Edge7000 IoT Gateway leverages an Intel Celeron processor to offer the performance needed to aggregate, verify, correlate, filter, and translate data at the edge. Another similarity is that both gateways are based on Intel IoT Gateway Technology, which provides a sophisticated, pre-integrated IoT software stack.

Figure 3. The Dedicated Computing Edge7000 IoT Gateway enables secure, seamless connectivity of a variety of devices.

The
Eurotech ReliaGATE 20-25 also demonstrates the merits of edge intelligence. This powerful Intel® Atom™ processor E3800-based IoT gateway is designed for industrial and rugged applications. Eurotech uses the gateway to run its Everyware Software Framework (ESF) for application development, as well as its Everyware Device Cloud (EDC) M2M/IoT device integration platform (Figure 4).

Figure 4. The Eurotech Everyware Device Cloud enables end-to-end intelligence.

Rolling Out a Smarter IoT

These are just a few examples of the ways developers can implement fog computing. For more examples of IoT gateways—many designed for specific use cases—check out the Solutions Directory.

Pattern-Matching: Low-Power Intelligence at the Edge

Posted on by Mark Scantlebury

One of the biggest challenges with deploying intelligence at the edge is keeping power consumption in check. Making smart decisions can take a lot of processing, which can burn a lot of power.

The Intel® Quark™ SE microcontroller C1000 tackles this dilemma with an innovative pattern-matching engine and a smart sensor subsystem. Figure 1 shows a diagram of the processor with the pattern-matching engine and sensor subsystem highlighted in orange.

Figure 1. The Intel® Quark™ SE microcontroller includes a pattern-matching engine and sensor subsystem.

With its new pattern-matching engine, the Intel Quark SE microcontroller can perform simple pattern detection on sensor data without engaging the main processor core. Inputs like vibration, temperature, or current can be passed directly from the sensor subsystem to the pattern-matching engine. If the engine finds a pattern it has been looking for, it can trigger a wakeup event to the main processor. The processor can then take whatever action is needed—such as starting up a piece of equipment or sending a message to maintenance staff.

The sensor subsystem also can operate independently of the main core. The heart of this system is a 32-bit ARC EM DSP core that can offload sensor processing from the main core. This allows the main core to stay in sleep mode until the sensor processing is complete.

The DSP core has hardware support for operations like multiply-accumulate, fractional arithmetic, floating point, divide, square root, and trigonometric functions. These arithmetic capabilities enable the core to perform initial signal processing and signal-conditioning functions like sensor fusion, data averaging, filtering, artifact rejection, and error correction.

Inside the Pattern-Matching Engine

While the sensor subsystem will be familiar to anyone who has programmed a DSP, the pattern-matching engine is more novel. The pattern-matching engine features 128 neurons that can perform two types of pattern recognition: K-nearest neighbor (KNN), where the input consists of the K-closest training examples, and radial basis function (RBF), which depends on the distance from the origin to correctly classify new instances.

Upon processing a pattern, the engine returns one of three states: identification, uncertain, or unknown. Up to 32,768 identification categories can be programmed.

When it comes to programming, neurons can have three states in the chain: IDLE, Ready To Learn (RTL), or Committed (see Figure 2). It becomes Committed as soon as it learns a pattern. A control line then changes its status from Ready To Learn to Committed. The next neuron in the chain then becomes the Ready To Learn neuron. The contents of Committed neurons, a representation of the knowledge they have built autonomously via learning examples, can be saved and restored.

Figure 2. Neurons exist in one of three states. (Source: General Vision)

Neurons are trained by example and decide autonomously when it is necessary to commit new neurons and/or correct existing ones. In the course of learning, only novelty—new information—results in a new committed neuron.

When a new example is presented for learning, the neural network first attempts to recognize it. If the example is recognized by one or more neurons, and they all agree on its category, the new example is discarded since it does not add any new information to the existing knowledge base. If the example is not recognized by any existing neurons, a new neuron is automatically added to the network to store the new example and its value.

Edge Intelligence in Action

One company taking advantages of these capabilities is Dublin-based Firmwave, which uses the processor in its Edge 3.0 wireless sensor platform (Figure 3). The Firmwave* Edge supports Zigbee*, Thread, RFID, NFC, Wi-Fi, Bluetooth*, LoRa*, SIGFOX*, and cellular connectivity. It comes with a host of on-board sensors, including temperature, humidity, light, pressure, position, and acceleration as well as a 36-pin expansion connector providing flexibility and extensibility.

Figure 3. Firmwave’s Edge wireless sensor platform is optimized for low power.

Adrian Burns, CTO and co-founder at Firmwave, notes that his company’s interest in Intel Quark SE microcontroller is in leveraging its on-board sensor subsystem to do data classification on the fly, offloading tasks from the core and enabling the processor to stay in low-power mode longer. He says, “We wake up the processor only when we want to do data connectivity to the cloud or gateway, allowing Firmwave Edge customers to reduce energy use.”

Burns notes that many of the company’s engineering efforts are centered on power optimization. When they see an opportunity to reduce power consumption, as the Intel Quark SE microcontroller does, “we have to be all over it,” he says.

A Smarter Way to Do Intelligence at the Edge

The Intel Quark SE microcontroller is in many ways a remarkable departure from Intel’s traditional approach. Adding the pattern-matching engine and DSP core in the sensor subsystem offload a considerable amount of processing from the main core, where Intel would traditionally focus its efforts. But it is clear that this new approach has resulted in high intelligence at low power. It will be interesting to see how the new capabilities will be deployed in the field.

Simplify Sensor Connections with Modular IoT Software

Posted on by Mark Scantlebury

One of the big challenges with IoT design is integrating all of the sensors. A typical IoT deployment has multiple sensors—such as temperature, humidity, and motion sensing—as well as multiple communication protocols—including Wi-Fi*, ZigBee*, and Bluetooth* low energy, to name a few. Sorting out all the firmware for the various sensors and networks can be a time-consuming and tedious process.

American Megatrends (AMI) is attacking this problem with tools that abstract the sensor interfaces. By providing a modular, plug-and-play approach, AMI says it can significantly speed up the early stages of the development cycle.

Modular Firmware

The foundation of the approach is the AMI RTOS* and its accompanying integrated development environment (IDE), shown in Figure 1. The IDE centers around schematics that show how sensor interfaces tie to pins of the system’s embedded controller. Developers need only select their preferred RF modules and sensors, and the dev environment will automatically produce customized firmware.

Figure 1. The AMI RTOS* is accompanied by an integrated development environment (IDE).

Because developers no longer need to learn the details of each interface, development time can be reduced significantly. In addition, this plug-and-play functionality makes it easier to swap components when design requirements change.

Sensor Hubs

The RTOS provides additional functionality needed to construct a sensor hub, such as a web-accessible device management user interface. Indeed, the RTOS is just one element in AMI’s larger offerings for IoT designs. As shown in Figure 2, AMI offers solutions that extend from the sensor hob to the IoT gateway and even into the cloud.

Figure 2. AMI’s scalable IoT solution architecture.

IoT Gateways

Devices running the RTOS can communicate with a gateway that runs AMI LINUX*. The OS is designed to enable efficient administration. For example, to add a new gateway or sensor array to the network, you scan a QR code into an app on a mobile device.

The gateway OS has multiple fail-safe mechanisms. If replacing hardware is necessary, getting operational again is straightforward with restoration of previous configurations, whether through a factory image restore option, image reinstall with a USB key, or directly from the cloud.

Efficient Edge Processing

Although AMI’s solutions can be used with any Intel® processor, the company sees particular advantages to using Intel® Quark™ processors. These processors have low energy needs as you would expect in embedded and IoT applications, and yet provide more performance per watt than other chips. That translates into important advantages in operations.

If sensor data needs significant filtering, an Intel Quark processor can handle that workload—along with some heuristics—at the sensor hub. Pushing this processing all the way out to the sensor hub frees the gateway up for other tasks. Hub-level filtering also reduces unnecessary data transmission, allowing gateways to communicate with more sensors than would otherwise be possible.

Cloud Connectivity

AMI LINUX provides connectivity to such popular cloud options as Amazon AWS*, Microsoft Azure*, and IBM Bluemix* to complete the data path. Developers can also use AMI CLOUD SERVICES, a private cloud service designed by AMI for IoT management and use.

AMI CLOUD offers significant administrative and operational advantages, like fail-safe backup for all sensor hubs and gateways. AMI CLOUD additionally protects uptime with a many-to-many rule engine that manages sensor rollover in case of a negative event. Robust data analytics close the loop so companies can make use of the collected sensor data and make results available for further data analysis and decision support.

Take the Pain out of IoT

In short, AMI’s software stack takes much of the pain out of IoT sensor integration—and it provides a convenient path to the cloud. To see other firmware and OS solutions from members of the Intel® Internet of Things Solutions Alliance, visit the Solutions Directory.

IoT Kit Puts Cloud Hooks in the Gateway

Posted on by Mark Scantlebury

The Internet of Things (IoT) is all about bringing the worlds of embedded and IT together. The question is how to unite these disparate technologies. Few companies have expertise in both embedded design and Big Data. Fewer still know how to combine these domains.

That’s why new offerings like the Advantech IoT Gateway Starter Kits* illustrated in Figure 1 are so important. These kits offer an end-to-end solution that makes it much easier to connect embedded devices to commercial cloud service like those from Microsoft, IBM, and Amazon. With these kits, companies can focus their design efforts on analytics and decision support rather than contentious configuration details.

Figure 1. The Advantech IoT Gateway Starter Kit in the context of an IoT application.

The Advantech approach begins with two ready-to-run industrial gateways—the ARK-1123H and ARK-2121L—based on Intel® IoT Gateway Technology and boasting a number of powerful features. But what makes the kits truly significant is the software that enables a smooth path from sensor to cloud.

Cloud Platform Built in

Most notably, the kits come with the WISE-PaaS/RMM cloud software, which can run on Microsoft Azure*, IBM Bluemix*, and Amazon Web Services* (AWS). The cloud platform provides a number of functions:

  • Data acquisition from the sensors
  • Remote control for the gateway system and I/O, dashboard console
  • More than 100 RESTful APIs, a popular type of standard web service for remote control and IoT application development
  • MQTT protocol support for communications to the cloud
  • Gateway monitoring to examine connected device and software conditions through a dashboard and system management console
  • Node-RED application logic designer

Node-RED deserves particular attention for companies that aren’t accustomed to designing and deploying embedded systems. A drag-and-drop interface allows engineers or IT professionals to design implementations without the need for embedded coding experience. Companies can quickly and easily build an IoT automation and data flow implementation.

Running on the gateway hardware is the WISE Agent software. It handles communications with the sensors, or the devices reporting up sensor results, whether through common serial I/O protocols, Wi-Fi, Ethernet, digital I/O, or USB, and bridges communications through the MQTT protocol to relay data to the cloud. The WISE Agent also responds to the RESTful APIs for remote control and IoT application development, so the cloud application built with Node RED, further customized in either C++ or C#, can fulfill sophisticated needs. The WISE Agent also consolidates the data and performs any processing requirements, so what the cloud receives is only what it needs.

The building-block development approach and automated communications and processing capabilities help companies undertake IoT development without having access to dedicated embedded development resources. In addition, professional technical support and access to a developer forum community help companies get the assistance they might need.

Powerful Hardware

Critical to the success of the IoT Gateway Starter Kits is the quad-core Intel® Celeron® processor J1900. The chip provides two major advantages to the kits. One is Intel® Active Management Technology, or Intel® AMT, which allows third-party management applications to better manage, repair, and control hardware, even when powered down. Even if a device is non-responsive at the OS level, software from the cloud can fix problems in the BIOS or restart the system.

In addition, the Intel Celeron J1900 processor comes with Intel McAfee Security Support*. Security is a major concern in IoT to keep hostile parties from interrupting activity or using an IoT device as a door into major systems.

Easier IoT Design

The Advantech kits are just one example of how members of the Intel® Internet of Things Solutions Alliance are simplifying IoT design. To see other IoT gateway solutions from members of the Alliance, visit the Solutions Directory.