- Advancing Technology Trends - Continued cloud computing technology growth may be limited with the growing need for faster real-time actions.
- Shifting to the Edge - Edge computing provides businesses more flexibility for IoT thanks to lowering costs and better performance efficiency.
- The Edge Computing Case for 5G - The next generation of high speed WWAN will enable innovation but needs new infrastructure and network topology.
- The Rugged Edge - As edge computing moves further away from the cloud, ruggedized platforms will be necessary to support new technologies.
- 5 Key Characteristics of Rugged Edge Computers
Advancing Technology Trends
The industrial world has gone under significant technological changes in the past decades through a digitalization of all its operations. The introduction of computers into traditional processes have accelerated innovation to help automate and optimize industries for efficient growth. In order to continue increased productivity and performance, new advents of complex smart sensors, artificial intelligence, big data, and advanced robotics combined with improved data communication is paving the way for a new industrial era.
Initially, the ubiquity of cloud computing was a primary driver for all new innovations by providing heightened performance infrastructure that could be accessed by any device with an internet connection. In fact, 73% of organizations have at least one application or a portion of their enterprise computing infrastructure in the cloud.1 Industries were able to realize greater cost reductions and maximize growth by migrating IT environments to the cloud.
This is also where big data established its dominance by generating new insights on business operations through incorporating smart Internet of Things (IoT) sensors and digitalizing machines to record every bit of data from the shop floor to the top floor creating valuable information for decision making.
However, industries today have quickly realized that a centralized cloud platform may not be a universal solution for all future applications to come. The sheer amount of data being generated from IoT sensors and devices cannot always funnel through a bandwidth pipeline to the distant cloud. Given current bandwidth limitations and network latency, applications that need actual real-time analysis or the ability to process larger complex datasets, face potential bottlenecks and would otherwise be unachievable. Aiming to break through this barrier for increased performance and at the same time provide immediate insights, data processing and communication needs to be located closer to the source, or also being termed the “edge.”
Shifting to the Edge
Moving away from the centralized aspect of cloud computing is where most new opportunities are starting to materialize. The rapid growth of advanced IoT will require more flexibility and performance in an effort to expedite applications in automation, machine-to-machine communication, and computer vision just to name a few. This can be achieved by integrating distributed edge computing platforms that are physically located closer to the source of data.
The “edge” in relation to a network layout sits right in-between the cloud and devices that generate data. While edge computing is not a new concept by any means, the advancement and proliferation of
smarter devices is transforming the way we interact with the world through digital platforms. The migration of data processing is shifting further away from the data center through these devices as more and more of them are interconnected through a shared peer-to-peer network.
Edge computing deployments are an ideal solution for a variety of circumstances. In contrast to having a collection of IoT devices connected to a central point on the cloud, an edge platform can offer better efficiency that doesn’t constantly siphon bandwidth and instead only transfer meaningful processed data.
For an example, consider a building’s smart security system utilizing motion-sensor IP cameras. In a strictly cloud based system, these cameras will continuously record video data and stream it to a server located elsewhere in a data center. The motion-sensor application reviews the recorded video data to only archive footage that featured motion activity. In this scenario, the security system puts a constant strain on the building’s internet infrastructure and substantial bandwidth gets consumed by the continuous stream of raw video data.
Alternatively, an intelligent edge computer or server can alleviate the amount of stress on the infrastructure by moving the motion-sensor computation onto a separate platform located in-between the camera system and the cloud network. The cameras will be connected to a nearby edge computer that analyzes all video data feeds for motion activity, and in turn, only transfers meaningful data packaged into a smaller size for archival on the cloud as needed. In this method, the security system minimizes network bandwidth and resources freeing up the server for other tasks as well as delivering significant cost savings.
In addition, an edge computer that handles all the local processing could in theory detach from the cloud network entirely giving businesses more flexibility in remote locations that might not have reliable internet connectivity. It could also work in a hybrid model that manages computational tasks at the edge for non-time sensitive data that can be compiled at the end of the day in the form of a daily report to a central cloud server.
The versatility of edge computers open up new opportunities for innovation given its ability to be deployed in varying environments. It is estimated by Gartner that connected edge devices will grow up to 25 billion units by 2021.2 Edge computing is a not a type of technology looking to satisfy market demands but a topology and framework that is being fulfilled with IoT technologies and cloud access to push forth new solutions. As the number of smart devices grow, a mediator in edge computing will facilitate the placement of proper processing hubs along the entire continuum that ranges from the core cloud at one end to the end device closer to source of data generation.
The progressive edge computing paradigm is made possible by several key factors bringing the concept into modern reality. The growth and enhancements of the platform is thanks to a range of smart sensors, actuators and interconnected devices each with their own memory, processors, storage and networking capabilities factored in with a trend of increased compatibility and ease of market access.
- Lower costs on compute and the Internet of Things
- Increased processing performance in smaller footprints
- Exponential growth of big data and bandwidth limitations
- Connectivity and interoperability between machines-to-machines
- Advancement in machine learning towards federated learning
Until recently, edge computing was not considered necessary due the strength of the cloud and its economies of scale. The centralized environment allowed data centers to maximize processing capabilities by operating at hyperscale and offered impressive computational proficiency. However, implementation realities must consider the limitations of bandwidth speeds that bottleneck applications needing real-time computing and minimal latency. Factored in with the economic costs of backhauling massive amounts of data and cloud ingress/egress expenses, enterprises are quickly realizing that both centralized cloud and distributed edge environments are necessary for future growth.
Much of the current attention on edge computing originate from the growing interest in developing IoT systems with the ability to deliver smarter data and valuable predictive insights. While edge computing affects nearly every corner of the IT landscape, the Industrial Internet of Things (IIoT) is at the forefront of the trend.
Information Technology (IT) and Operational Technology (OT) are converging in numerous industry sectors including healthcare, transportation, manufacturing, defense, aviation, mining, energy, utilities and telecommunications. Increasingly, connected systems featuring wireless sensor and actuator networks are being integrated into the management of industrial environments, such as those for water treatment, electric power and factories. The combination of automation, communications and networking is an integral part of the growing IIoT. Edge computing provides the platform capacity to efficiently filter and analyze the large quantities of data generated from IIoT devices for local real-time actions.
The modernization of operations through innovative IT solutions enables more direct control and cohesive monitoring to infer more value from machine-driven data. As the adoption of IIoT steadily rises, industrial organizations are finding ways to help them further monetize the value of their assets and achieve meaningful results.
The Edge Computing Case for 5G
One prime example that will be the result from the culmination of edge computing is none other than 5G networks, the fifth generation of cellular network technology. Just like its predecessor, 4G, the track towards 5G was developed in direct response to the continued exponential growth of smart devices needing improved internet network connections to manage large amounts of data. In comparison, 4G broadband can only support up to 2,000 active devices in a square kilometer while 5G standards are designed to support up to 1 million connected devices in the same space paving the way for massive IoT networks.3
Processing higher volumes of data at speeds 10 times faster compared to 4G will require an entirely new landscape of systems starting from the network infrastructure all the way to every single connected device. The evolution to 5G will drive a necessity for edge computing platforms given the complexities of real-time inference for massive IoT density. The reason being that 5G works with extremely high bandwidth frequencies that allow for faster speeds but reduces overall range, providing a structure for real-time processing power.
So in order to distribute 5G capabilities reasonably, small cell base stations need to be integrated throughout dense locations to provide ample 5G coverage. This is exactly what defines edge computing, a system that stands as close to the end-user as possible to deliver faster processing speeds to support real-time computing. 5G technology will have to rely on edge computing hubs deployed at each base station to enable time-critical applications such as autonomous vehicles, high-frequency trading, augmented reality and industrial automation.
These emerging technologies require massive amounts of real-time computation to deliver content or instructions to users with reliability and availability anywhere at any given time. This digital transformation will help reduce back-haul traffic by keeping data content at the edge and provide increased quality of service. Edge computing plays a pivotal role in the arrival of 5G, solving a variety of challenges for latency, management, security and monitoring.
However, some obstacles lay in the way for edge computers to fulfill the need for a successful IIoT and 5G integration. When implemented in rigorous industrial environments, sensitive computer components face a risk of breaking down under extreme conditions where stability and continued operability are the standard for traditional machinery. For edge computers to thrive, industrial ruggedization is necessary to deliver new computing functionality in situations where typical computers are unsustainable.
The Rugged Edge
Often referred to as Industry 4.0 or the Fourth Industrial Revolution, the digitalization of heavy industry is a larger than life transformation of equipment and processes to build an end-to-end ecosystem that leverages the advantages of edge computing and IIoT. This ecosystem involves every facet throughout enterprise resource planning (ERP) from internal functions including sales, procurement, engineering and R&D, all the way to external players in suppliers, logistics and end-users.
The application topology of edge computing moving data, applications, service and power closer to the source of data generation can attain valuable insights for industrial factories looking to make better informed strategic decisions. The benefits of digitalization does not only stop at increasing production efficiency or expanding tactical regionalization, it also opens the possibilities for developing new products and technology for new untapped verticals.
The benefits of incorporating edge computing are clear, however, when deployed in real-world environments that are harmful to electronic components, a specialized rugged industrial computer is required to bring mobile edge computing capabilities to new frontiers.
For instance, heavy machinery used in deep underground mining operations are exploring the integration of smart IIoT sensors and rugged edge computers to enable remote control and driverless equipment. Human operators can safely control heavy drilling jumbos and loaders to further increase safety, productivity, and efficiency. Given the nature of underground mining, any sensitive computer component would be severely affected by the high amount of shock, vibration, hot temperatures and dust. To combat this issue, the edge computer needs to be fortified and validated to operate reliably under strict conditions. The addition of IoT data and edge computing also provides the mining industry better insight through process optimization to boost asset utilization, predictive maintenance, energy management and real-time analytics for advanced modelling when making strategic decisions.
Rugged edge computers are specifically developed to withstand the rigors of harsh usage conditions such as the industrial mining example and are able to achieve a high level of durability through incorporating ruggedized features throughout its entire product design. From the external enclosure to the internal components, every piece of a rugged edge computer is purpose-built through a combination of mechanical and thermal engineering to address the issues of strong vibrations, extreme temperatures and wet or dusty situations.
Rugged Edge Computer Key Characteristics:
- Wireless Connectivity - Access to 4G LTE and future 5G networks including GPS/GNSS location tracking allows more flexibility and possibilities for devices to gather important data points for new technologies. In addition edge computers can process local data while transferring and offloading more compute intensive tasks to a cloud server thus increasing overall efficiency.
Mobility and Remote Deployment - A key feature for rugged edge computers is the capability to be deployed in various mobile and remote locations decentralized from resources in the cloud. These type of applications often times have space fitting constraints and require compact form factors to fit into tight-spaces. Therefore, a key foundation for mobility and remote deployments heavily rely on mission critical capabilities that ensure 24/7 reliability due to the lack of on-site technical support. The idea for mobility and machines actively running in remote locations leverage “watchdog” functions that allow for remote self-reset and system monitoring from a hub. Even operating software layers (Win 10 IoT) are designed to support edge and IoT functions with long-term service channel options, which basically avoid updates that can cause interruptions and detrimental downtime. One example of edge computing is for in-vehicle environments , expanding the reach of computing power where current integrations are less capable or more decentralized. Access to CAN bus protocol enables predictive maintenance, one of the most sought after applications for railway, fleet managers, heavy machinery operators and first response vehicles. With deeper and real-time insights to vehicle's engine conditions, businesses can greatly reduce costs and downtime through scheduling repairs based on accurate data.
Another in-vehicle feature that is best for rugged edge computers is power-ignition management; once toggled in car mode versus pc mode, this allows the computer to communicate directly with the car’s ignition and controls the on/off delay to ensure applications are not interrupted during the ignition process. Learn more about power ignition management.
- Performance Accelerators - Gone are the days of simplistic edge devices that merely handed off data information from IoT sensors and actuators to the next stage for actual computing. Increase in processor performance, GPU or VPU based inference analysis, and large capacity data drives have empowered rugged edge computers to manage local analytics without having to move data over the network to the cloud thus saving bandwidth and eliminating latency effects. This opens the realm of real-time computing and better application performance. Autonomous robots can locally process camera information and immediately act upon the task at hand such as stopping its operation when something is blocking its way. Having to send camera information over a network for processing not only risks slow reaction times but runs into the issue of rising bandwidth costs and unstable connections.
Variety of I/O Ports and PCIe/PCI Expansion Slots - What allows edge computers to work so well is the capability to connect to various legacy and modern devices for data gathering both in legacy analog and digital IoT signals. By including GPIO and Digital I/Os, edge computers can capture signals from sensors of all kinds for actionable results. The ability to support multiple Universal Serial Buses (USB 2.0 and 3.0) are also key in today’s digital data transformation framework due to these sensors and devices leveraging high-speed and superspeed transfer rates up to 10 Gbps. LAN/PoE port availability also enable smart cameras that are advancing at a rapid pace for better depth perception and image processing for computer vision technology in digital surveillance. In addition to that, rugged edge computers also have lockable M12 connectors for most I/O ports that secure the system in cases of environments where shock and vibration are present and can endanger operations. Learn more about the most popular industrial, embedded and edge computing I/O ports.
Another critical feature is to have flexibility for application specific performance add-in cards through common computing standards like legacy PCI or current PCIe Gen 2,3 and now recently 4. This flexibility provides not only fast data transmission speeds but options for applications specific performance cards like GPU, VPU, or capture cards that communicate through the PCIe bus protocol.
Ruggedness and Security - Without proper protections against external elements and factors that are harmful to sensitive computer components then edge computing won't be able to provide the benefits of real-time analytics in harsh operating environments. Features include fanless designs, wide operating temperatures, varied voltage input ranges and high shock and vibration reliability. Industry 4.0 markets looking to integrate better data management will require rugged edge computers to unleash the next wave of innovation and growth. In addition to rugged environmental safety is the need for better security measures against potential risks and attacks. Edge computers configured as a nearby security agent for new walls of defenses in between IoT devices and outside connections.
In addition, at the core of edge computing its imperative that deployment models ensure strengthened security, accelerated performance, and increased scalability. Security layers are equally important on both hardware and software layers, each working together to provide authentications for critical operations in mobile and remote locations. For example the Trusted Computing Group publishes the Trusted Platform Module (TPM 2.0) hardware specification that is an international standard that adheres to modern security and privacy protection features. According to Trusted Computing Groups press release, “[TPM 2.0] implementation in computing devices safeguards cryptographic keys; prevents private keys from being exported; shields PIN values used for authentication; and records and anonymously reports on software loaded during the boot process to protect against malware and attacks. As such, it will become an essential component of any comprehensive security strategy”… “The specification was developed as a set of commands that enable the TPM to be used in a variety of systems and devices, from PCs, servers and networking gear to embedded systems and the Internet of Things.”4
Industry 4.0 markets looking to integrate better data management will require rugged edge computers to unleash the next wave of innovation and growth. In addition to rugged environmental safety is the need for better security measures against potential risks and attacks. Edge computers configured as a nearby security agent for new walls of defenses in between IoT devices and outside connections.
2. The Edge Completes the Cloud — A Gartner Trend Insight Report [↩]
3. 5G - Connection Density — Massive IoT and So Much More [↩]
4. Trusted Computing Group TPM 2.0 Library Specification Approved as an ISO/IEC International Standard Date Published: June 29, 2015 [↩]