#edge computing for iot

Mobodexter | Mobodexter

Kubernetes Edge Computing: Chick-fil-A, known for its addictive poultry sandwiches and also waffle french fries, is reported to become the third-largest U.S. fast-food chain behind McDonald’s and Starbucks. Behind the scenes, their business remains at the forefront of embracing a powerful modern technology like Edge computing and Kubernetes.

Chick-fil-A published a Medium post that it will undoubtedly be running Kubernetes on the Edge of 6,000 IoT devices in all 2,000 of its dining restaurants. Part of the chain’s internet of things (IoT) strategy to gather and examine more information to enhance customer service and also functional effectiveness. For example, they can forecast how many waffle french fries need to be cooked every minute of the day.

This case study shows why Kubernetes has quickly ended up being a crucial active ingredient in edge computing. A tried and tested and effective runtime system to help address particular challenges across telecommunications, media, transportation, logistics, farming, retail, and also other market sections.

The telco industry particularly has much to acquire from edge computing. As competitors among operators intensify, telco firms must distinguish themselves with new use-cases such as commercial automation, virtual reality, connected cars and trucks, sensor networks, and smart cities. Telcos significantly are using edge computing to ensure these applications function flawlessly while also driving down the prices of deploying and handling the network framework.

Just what is edge computing? 

Edge computing is a variation of cloud computing, with your infrastructure like compute, storage, as well as networking physically closer to the devices that create information. Edge computing enables you to position applications and solutions closer to the source of the data. This placement gives you the twin advantage of reduced latency as well as lower web traffic. Reduced latency boosts the performance of field devices by enabling them not just to react quicker, yet to also respond to even more events. Lowering web traffic helps in reducing prices as well as boosts overall throughput. Whether an application or service will be in the edge cloud or the core datacenter will rely on the specific use-case.

Intel Edge Computing Framework

How can you build an Edge cloud? 

Edge clouds need to be built with the very least two layers. Both layers will certainly maximize operational efficiency as well as designer performance, and each layer is created differently.

The initial layer is the Infrastructure-as-a-Service (IaaS) layer. In addition to giving compute and storage space sources, the IaaS layer needs to address the network performance needs of ultra-low latency and also high data transfer. IaaS is the layer where blade systems from HP, Dell, IBM, and Lenovo, and specialized systems like Lenovo Think system SE350 come in the picture.

The 2nd layer is the Kubernetes layer, which gives an ideal platform to run your applications and services. Whereas making use of Kubernetes for this layer is optional, it has confirmed to be a reliable system for those organizations leveraging edge computers today. You can release Kubernetes to field tools, edge clouds, core data centers, and the general public cloud. This multi-cloud implementation ability provides you total flexibility to deploy your applications anywhere in the field, Edge, or cloud. Kubernetes deals with your developers the capability to simplify their DevOps practices and minimize time spent integrating with heterogeneous operating environments.

Benefits of Kubernetes on Edge

In addition to very low downtime and outstanding performance, Kubernetes offers numerous built-in advantages to resolve Edge compute obstacles, including versatility, resource performance, performance and reliability, scalability, and observability.

1) Flexibility

Kubernetes decreases the complexities associated with running computing power across various geographically dispersed points of existence and varied architecture by supplying versatile tooling that allows designers to connect with the Edge seamlessly.

With Kubernetes behind our Edge platform, users can run application containers at scale through an agnostic network of distributed computing infrastructure, which in turn, extends complete flexibility to the users to be able to run the application anywhere along with the Edge computing.

2) Resource Efficiency

Containers are light-weight by nature and enable you to use the underlying infrastructure in an extremely effective method. However, managing thousands or, in many cases, countless containers throughout a distributed architecture gets complex very rapidly. Kubernetes supplies the underlying tools to effectively manage container workflows through automated networking, storage, and event logs.

The Kubernetes Horizontal Pod Autoscaler is one key feature that naturally provides itself to edge computing performances. It immediately scales the variety of pods up or down in a replication controller, release, or replica set based upon latency or volume thresholds. Think about a point of existence in an edge location that requires to manage abrupt traffic increases, like, in the case of a local sporting event. Kubernetes can auto-detect traffic from event logs and provide resources to scale to the fluctuating demand. This kind of auto-scaling takes the uncertainty out of forecasting and preparing for infrastructure needs. It also makes sure that you’re only provisioning for what your application requires in any given period.

3) Performance

Modern applications require lower latency than standard cloud computing designs can provide. By running workloads better to end-users, applications can recuperate vital milliseconds to give a much better user experience. As mentioned above, Kubernetes’ capability to respond to latency and volume limits (utilizing the Horizontal Pod Autoscaler) implies that traffic can be routed to the most optimal edge places to reduce latency.

4) Dependability & Scalability

Among the significant advantages of Kubernetes is that it is self-healing. In Kubernetes, you can restart containers that stop working, replace, and reschedule containers when nodes fail and eliminate containers that do not respond to your user-defined health check.

 In Kubernetes, services abstract this process, allows you to begin and stop containers behind the service. At the same time, Kubernetes carries on managing traffic reroutes to the right containers to avoid service disturbances.

Besides, since Kubernetes’ control plane can handle tens of countless containers running over numerous nodes, it allows applications to scale as needed, especially fitting the management of distributed edge workloads.

5) Observability

Knowing where and how to run edge work to make the most of efficiency, security, and performance needs observability. However, observability in a microservices architecture is complex.

Kubernetes offers a full view of production work, enabling optimization of efficiency and efficiency. The built-in monitoring system allows real-time insights (consisting of transaction traces, logging, and aggregated metrics) with instant provisioning figured out by configuration settings. 

Another point of observability is of specific interest when it concerns edge computing is dispersed tracing. Distributed tracing permits you to collect and construct an extensive view of requests throughout the chain of API calls made, all the way from user requests to interactions in between numerous services. With this info, you can determine traffic bottlenecks and opportunities for optimization.

Interest in edge computing is being driven by rapid data increases from smart tools in the IoT, the coming influence of 5G networks as well as the growing importance of performing artificial intelligence jobs at the Edge. All of which need the ability to deal with flexible needs as well as shifting workloads. Kubernetes Edge Computing is one of the rapidly growing technology that companies like Mobodexter are continually innovating and help customers to become productive and efficient with their IoT strategy. 

Footnotes:

Mobodexter, Inc., based in Redmond- WA, builds internet of things solutions for enterprise applications with Highly scalable Kubernetes Edge Clusters that works seamlessly with AWS IoT, Azure IoT & Google Cloud IoT.

Want to build your Kubernetes Edge Solution – Email us at [email protected]

Check our Edge Marketplace for our Edge Innovation. 

Join our newly launched marketing partner affiliate program to earn a commission here.

We publish weekly blogs on IoT & Edge Computing: Read all our blogs or subscribe to get our blogs in your Emails. 

Mobodexter | Mobodexter

Real-World Examples of IOT Edge: IOT Edge computing is a framework that brings computing in the internet of things closer to the devices, which are the endpoint of the IOT network. By bringing the data processing closer to the Edge, it accelerates the processing, reduces the time taken for making decisions, and thus drives real-time action. IOT edge computing has gained enormous frenzy in recent years. Different versions of edge computing like fog computing have also gained momentum. Various applications, solutions, hardware, software have come into the market to enhance and propagate the adoption of edge computing among the IOT implementers. There are going be to be several advantages for the early adopters of the technology. Edge computing has also advanced to include AI and machine learning customized such that it is adaptable to edge computing. One of the significant benefits of Edge computing is the reduction of latency.

There are several early movers with IOT edge adoption. We will look at some of the real-world examples of where edge computing has been implemented and where it is bringing in operational efficiency.

Indoor Farming

Vertical Indoor farming has been an exciting field that benefits significantly from IOT. In indoor agriculture, the plants are mounted with sensors that track the water needed for the plants and the environmental conditions of the plants. The data from these sensors helps to adjust the operating conditions for these plants to achieve the maximum nutrients, flavor, and avoid wastages. These data also provide information when the vegetables can be picked and thus avoid wastage. Building an edge solution in this environment helps in the timelines of the decision and improve the growing environment.

Autonomous Vehicles 

Autonomous Vehicles is another interesting domain that can immensely benefit from edge IOT computing. Driverless vehicles need to analyze much information regarding their surroundings and communicate with the other cars on the road. An autonomous vehicle can send a large volume of data from all its sensors for it to perform optimally and efficiently. And as more and more self-driving cars are introduced, the data being sent by these sensors from all cars could put an enormous strain on the network bandwidth. If all these data needs to go to the cloud for processing, it will be killing the network. So these self-driving cars will benefit significantly if these have edge computing implemented. Some of the decisions by the vehicles can be local if the processing and AI are locally available.

Healthcare

The healthcare industry is another use case that can immensely benefit from edge implementation. It helps to localize the data processing for patients in the hospitals. Rural areas would be most beneficial in diagnosing and administering the action. 

Industrial Automation

Many factories are now adopting the internet of things, and edge computing will be one of the enablers for smart manufacturing. Sensors implemented in the factory gather data on the machine and factory conditions. Localized processing enable action on these data in real-time. For example, predicting a machine failure in real-time can help in taking corrective action to fix the machine in time and thus save a lot of time and cost. Similarly, data gathered can also help to determine the quality of the product in the factory and can drive any fine-tuning necessary to make changes in real-time.

Smart City

A smart city project includes traffic management, smart buildings, smart parking, smart streetlights. Data from a smart city could be voluminous. Sending this amount of data to the cloud could strain the bandwidth of the network, and getting a real-time response for the data could be not as quick as needed. Implementing edge computing can help some of these entities to make an autonomous decision and in real-time.

Real-World Examples of IOT Edge: When large amounts of data needed to process in a time-critical way, Edge computing would be the way to go. Edge computing can handle a large volume of data and improves the response to events by processing the sensor data locally. This process removes the need to send a large amount of data to the cloud and reduces the time to respond to the event. Also, lowering the cost involved in storing large volumes of data in the cloud. It also reduces the network bandwidth utilized in sending large amounts of data to the cloud. It also helps in securing the data by maintaining the data locally.

Footnotes:

Mobodexter, Inc., based in Redmond- WA, builds Internet of things solutions for enterprise applications with Highly scalable Kubernetes Clusters on the Edge. 

Check our Edge Marketplace for our Edge Innovation. 

We publish weekly blogs on IoT & Edge Computing: Read all our other blogs or subscribe to get our blogs in your Emails. 

Mobodexter | Mobodexter

New Features in Raspberry Pi 4: Raspberry Pi, the learning computing kit, became a standard in low-cost Single Board Computing (SBC). Since the launch, Raspberry Pi has been powering everything from robots, smart home devices, and even digital kiosks. The new Raspberry Pi 4 takes Pi to another level. Its performance is that’s good enough to use as a desktop P.C. with the ability to output video at 4K 60 Hz resolution or use with dual monitors. Hence it can be used for more complex applications. 

Major Differences

The most important new features are :

a) the faster processor and GPU.

b) the faster RAM.

c) the addition of USB 3 ports.

d) Dual micro HDMI ports and 

e) Support for 4K HDMI output. 

On-board Ethernet port run like a real Gigabit connection with 125 MBps. The microSD card slot is also 2X as fast, offering a theoretical maximum speed of 50 MBps versus 25 MBps on the 3B+.

Because the new processor needs more power, the Raspberry Pi 4 B charges over USB Type-C unlike micro USB used by its predecessors. It also requires a newer power adapter that can deliver at least three amps of power and 5 volts. USB Type-C connectors are useful as they are reversible and more comfortable to plug in.

New CPU & Memory

The Raspberry Pi 4 has similar dimensions and design to all previous models. However, it is an all-new platform, powered by a new processor from Broadcom BCM2711B0. Since the first Pi launch in 2012, all Pis have used 40nm SoCs. BCM2711B0 is a new chip based on a 28nm process. It uses ARM Cortex-A72 instead of the older Cortex-A53 microarchitecture. The BCM2711B0 chip has four cores and is clocked at 1.5 GHz. The ARM Cortex-A72 processors are significantly faster and use more power than their A53-powered ancestors.

The RAM is also a bit quicker. Pi 3B+ had 1GB of DDR2 RAM. On the Pi 4 B, it goes up to 4GB of DDR4 RAM. In addition to the increased bandwidth, the additional memory option is a huge deal, particularly for web surfing.

New Video Processor – GPU

The GPU got a nice performance boost too. GPU moves from a Broadcom VideoCore IV that operated at a clock speed of 400 MHz to a VideoCore VI that operates at 500 MHz. The new GPU allows it to output to a display at up to 4K resolution with a refresh rate of 60 fps or to support dual monitors at up to 4K resolutions with a refresh rate of 30 Hz.

Storage

Storage is critical, no matter how fast are your processor, RAM, and GPU. If your storage is slow in an SBC, everyday tasks like opening apps and files will be slow. Raspberry Pi 4 B’s primary storage device is its microSD card reader. MicroSD is convenient but a bit constrained. The Raspberry Pi 4 B has a top transfer rate of 50 MBps, and that is double the speed of the card reader on the 3 B+.

A speedy USB drive or an external SSD can get you a far better storage performance out of the Raspberry Pi 4 B. Also the Pi 4 B comes with USB 3 ports of speed at 625 MBps.

GPIO Pins

The great thing on any Raspberry Pi is the GPIO (General Input / Output) at 40 pins. The pin count and pin layout remain unchanged from all the prior models. Hence any “hats,” sensors or LED screens that were made to attach to a Pi 2 or 3 will be compatible.

Networking

The Raspberry Pi 4 supports 802.11ac Wi-Fi as its predecessor. However, it has Bluetooth 5.0 support, an improvement over the Bluetooth 4 on prior models. The Ethernet port now has more bandwidth, that allows it to offer a full gigabit throughput while previous models could only achieve about 330 megabits.

Power and Heat

Pi 4 has A more power-hungry processor and the need for at least a 5-volt, 3-amp power adapter. Hence it is be expected to consume more power than its predecessors. Under load, that power number jumps to 7.6 watts, but that’s still only 19 percent more juice than its direct predecessor.

A.I. and Machine Learning

The most compelling new use case for the Raspberry Pi 4 is for A.I. & machine learning. With the earlier Pis, you could use a camera to do simple object detection at low frame rates. However, the added performance and I/O from this new model should open up a whole new world of use cases.

With an optimized A.I. framework, Pi 4 should be able to do real-time facial and object recognition. Moreover, because it has USB 3, an A.I. accelerator like the Google TPU USB dongle will have much more bandwidth to send data back to the SoC.

Conclusion

New Features in Raspberry Pi 4: The Raspberry Pi 4 represents a significant improvement, not only for the Raspberry Pi but for single-board computing. For the first time, it is realistic to use your Pi as a secondary P.C. However, the more substantial real benefit will come not from folks who use Raspberry Pi 4s instead of x86 P.C. Also from all ll the innovators who harness the system’s enhanced performance, I/O and graphics to create new IOT devices, media servers, and robots. 

Try Raspberry Pi 4 to enjoy these New Features in Raspberry Pi 4. Available now at a special 20% discount on our Edge Marketplace here. 

Footnotes:

Mobodexter, Inc based in Redmond, WA builds Internet of things Edge solutions for enterprises applications on Kubernetes Edge Clusters. 

Check our Edge Marketplace. Special 6th Anniversary Discount sale of 20% on all products. 

IoT Developers can try Edge Platform here. 

Mobodexter’s IoT Edge Blogs: Read

Mobodexter | Mobodexter

Mission-critical Edge: Internet of things is growing at phenomenal phase. The nucleus of IOT is the data, and as IOT is expanding, so is the data. With the growth in IOT, a sensor is attached to any, and all devices. These sensors keep churning out the data every second; what we are talking about is trillions of Gigabytes of data generated by IOT devices. Cisco predicts that by 2020, IOT will generate 600 trillion GB of data. 

These massive volumes of data will drive data management challenges for IOT. In the current architecture, the data collected by these sensors are all moved to the cloud. For example, in the case of traffic management, the data from different sensors across the city are being routed to multiple clouds sometimes to partner cloud. The data in the cloud are analyzed, and some anticipated events are driven based on the processing. The real potential of IOT is when these data can be used to create real-time actions. For instance, in the case of traffic accidents, the data collected by street sensors need to be processed in real-time. After that, only it can give live feedback in re-routing the traffic in avoiding congestions. In these scenarios does it mandate to loop the data to the cloud sometimes to partner cloud and back. The real value will be to process the data, analyze, and drive action at the Edge for quicker response time. In situations where the data changes spontaneously warrant an instinctual response. 

So the data management in mission-critical IOT application becomes crucial. Moreover, Edge is how data management can be made efficient in IOT mission-critical or not. Let’s take an example of non-mission critical use, an IOT smart home device like a thermostat does not need to send the data to the cloud every time. Edge has to make intelligent, the intuitive decision on data routing to the cloud. Edge has to determine whether collected data needs a real-time response. Hence data will have to be crunched immediately or is more of a storage and pattern collection data, thereby sending to the cloud. Also, if it is routing the data to the cloud where and which database does it go. Since only the necessary data travels outside the edge network, it is less likely to get into the hands of the wrong person.

So it is clear that for the best approach in edge-cloud architecture IOT data management tasks should be handled by Edge.

When talking about data management, we need to understand the type of data that is produced at the Edge. First this the data which requires a real-time action so let’s call these real-time data, these needs to crunched instantly and action-driven immediately. The second type of information is the data which needs to be stored in the Edge for a short time, and then the data can be sent in batch to cloud for later analysis.

For Edge to support data management, it should support storage which means it should support database. These databases have to be specifically designed for edge scenarios since these have to be efficient and will be built in a constrained architecture. Building data governance at the Edge helps in deciding which data needs to be stored and where it needs to stored and thus saving costs on storage. Building AI on the Edge helps to drive intelligent decision making at the Edge in real-time. However, at the same time, to better utilize the constrained environment, it is better to have the analytical engine on the cloud.

Mission-critical Edge: We are reaching a stage in the IOT and Edge revolution. In this revolution, data from the Edge to the cloud is refined and enriched. A robust working architecture of edge-cloud helps in improving the efficiency of the whole IOT platform.

Footnotes:

Mobodexter, Inc based in Redmond, WA builds IoT Edge solutions for enterprises applications on Kubernetes & Dockers.

Register now at developers.paasmer.co and try a free Edge trial license of Edge Intelligence software that runs on Raspberry Pi.

Follow our all our weekly IoT Blogs: https://blogs.mobodexter.com

Our IoT Newsletter reaches 3000+ subscribers– Subscribe Now.

Become our affiliate partner today.

Mobodexter | Mobodexter

Edge Computing for Industry 4.0: Organizations are starting to understand that they require unwavering control over every element of their organization to drive digital change. How they do this? At the edge.

Up until edge computing is as simple to establish and handle as a consumer gadget like a video games console. However, it will not be available to a lot of the commercial applications that it is most advantageous.

What do we mean by this?

As Industry 4.0 emerges, control and management of interlinked innovations (IIoT) at the application layer (edge) are critical. It allows the commercial business to take actions towards real-time and proactive management of applications, instead of the ad-hoc reactive control environment of market 3.0. With edge computing comes the ability for commercial applications to handle information where it is typically most helpful when it is produced. It enables analytics at the point of requirement, at a time and in a location where it can be actioned. It opens brand-new levels of control, effectiveness, and performance for market and genuine ROI for the interlinked business.

Evaluating the increasing quantity information, produced from an entire range of sources, can yield insights that even the most skilled professionals would not have even imagined. That’s what makes it so exciting.

When a company has an alignment on their objectives, they then gather the information and utilize analytics to take a look at it. Getting information from the application edge is the sharpest of mining tools, so having edge computing opens a world of chance.

Computing at this level, nevertheless, suggests that a brand-new, vitally interlinked application edge environment is running on the plant floor. Edge environment is at the Operational Technology (OT) level, along with in the server room and cloud environments at the IT level.

This blurring of the lines in between OT and IT innovations is part and parcel of a merging in changes that are opening the possibilities of market 4.0, however it is not without useful ramifications. Today, it’s much more difficult to release and manage software application at the edge. Because it remains in an information center or a cloud where the IT systems for the security, servers, networking exist. To release at the edge needs workarounds that include developing a series of mini-data-centers with associated abilities at the edge.

This layer of IT facilities requirement required at the edge will keep back its implementation unless it can be made much more comfortable with producing. If the OT engineers are releasing these mini-data centers all around the factory, then how can the IT department maintain and keep the levels of cybersecurity and networking abilities to support them? The response is that they can’t.

This observation of real-world applications of edge computing drives the design of many companies edge products and platforms.

Control of data

By providing a regular schedule of their server platforms, consumers can count on the stability and efficiency of the information produced. This approach marks an essential shift for market. It is one that unlocks to digital transformation and permits them to take the next actions towards more self-governing choice making. With reliable data returning from the edge environment, you get to the next level of analytics, then the insights and after that cycle back to manage. Eventually, what companies desire is machines and systems that drive themselves, that’s the AI part of it.

Conclusion:

Edge Computing for Industry 4.0: The market is beginning to take control of the data that is being produced in the period of IIoT. Moreover, in looking for to manage the information, to gain from it, commercial leaders understand that keeping a significant part of it in the context and environment of the application itself is better. Analyzing data in the edge environment naturally suggests that calculating requirements to occur there, and Edge computing is providing this ability. Notably, the business is also making all of this not merely possible, however also useful. Edge computing is making it simple, helping the factories to strive towards the autonomy they seek. However, it is also building in the security and networking functionality that reduces the IT implications of computing in the OT environment.

Footnotes:

Mobodexter, Inc based in Redmond, WA builds IoT Edge solutions for enterprises applications on Kubernetes & Dockers.

Register now at developers.paasmer.co and try a free Edge trial license of Edge Intelligence software that runs on Raspberry Pi.

Follow our all our weekly IoT Blogs: https://blogs.mobodexter.com

Our IoT Newsletter reaches 3000+ subscribers– Subscribe Now.

Become our affiliate partner today.

Mobodexter | Mobodexter

Edge Computing Layers: Edge computing will create a transformational impact on American society and business, thereby enabling new technologies and services across instant wireless connectivity.  For that to happen, edge computing must navigate a confusing maze and conflicting definitions, that are caused by the diverse nature of edge infrastructure and business models.

Significant stakeholders in edge computing are taking steps to clarify the various infrastructure layers and use cases. Multiple associations are being created by large corporations to define edge computing standards. Thes steps are essential to lay the groundwork for success. Edge computing is not a single entity like a computer, but several technologies that will work together to create a solution to solve a particular problem.

However, there are many challenges. Ege computing implementation timing remains unclear with many enterprises on when the business demand for the different flavors of edge infrastructure will materialize. This uncertainty is complicating decisions on how and when to fund these projects. 

Edge needs new technology:

Edge computing is not a single technology. It is instead a phrase that describes several layers of infrastructure, some of which are updates of several current models like a high-speed cellular connection – 5G. Edge computing solution is a response to new technologies such as autonomous vehicles and distributed AI applications which need low latency and proximity to users. These technologies are transforming the future of the Internet.

The transformation to edge computing will be an enormous effort for enterprises. It will take many years, cost billions and billions of dollars, and ultimately become massive in scale.

Multiple Layers and definitions:

For edge computing to be deployed at a vast scale, the edge will need to be compact, cheaper, and take the network to other new places. The critical components of edge computing include wireless networks, mobile devices, telecom towers, small cells, and distributed antenna systems as well as data centers and cloud platforms. Edge is a universe that transforms traditional business models built on specialization.

The data center industry defines an edge is a module. The small cell industry defines an edge is a small cell. Device makers envision analytics will run on sensors.

For cloud computing, it will take time for the market to grasp a holistic understanding of the edge and its many layers. Industry veterans believe things have progressed further due in part to the scope of the business opportunity.

Five layers of infrastructure:

Data Sources: Wearables, Things & Sensors

Processing Endpoints: Cars, Robots or Personal Computers

Local Data Hubs: Data processing at a small IT room or a building

Urban Data Hubs: Data processing at a cell tower or metro data center.

Regional Data Hubs: Cloud availability zones and colocation center.

The Linux Foundation offers a chart of edge layers and provides four principles that frame the participants’ lens on edge:

The edge is a location and not a single thing.

There are many edges, but the edge of the last mile network is essential.

The edge has two sides: an infrastructure edge and a device edge.

Data Computing will happen on both sides, with edge working in coordination with the centralized cloud.

Disruption demands new solutions:

This revolution is going to cost to implement. Many executives predict that the buildout of edge computing will pressure the economics of data infrastructure. Users will not pay a premium to go faster. Demand for services has a way of doing economics work, even in cases like this where the disruption is along the way.

There will be data centers in these new places, but they may be built and powered very differently than the server farms of Silicon Valley and Northern Virginia. The core compute doesn’t exist in one place like a datacenter anymore. It is instead distributed across multiple layers and locations. Edge Computing Layers are essential for real-world implementation, and they need careful strategic planning to create a transformational business impact.

Footnotes:

Mobodexter, Inc based in Redmond, WA builds IOT Edge solutions for enterprises applications on Kubernetes & Dockers.

Register now at developers.paasmer.co and enjoy free Edge trial license of Edge Intelligence software that runs on Raspberry Pi.

Follow our all our weekly IoT Blogs: https://blogs.mobodexter.com

Our IoT Newsletter reaches 3000+ subscribers– Subscribe Now.

Become our affiliate partner today.

Mobodexter | Mobodexter

Edge Computing for IIoT: Companies are racing to adopt the Industrial Internet of Things (IIoT) technologies. Many are evaluating edge computing for efficiencies, where data from IoT devices is processed on-premise, rather than in the cloud. The increased processing speed of edge computing allows businesses to analyze data and make critical decisions in almost real-time, thereby drastically drive to improve operational efficiencies.

However, introducing these new devices and endpoints also introduces new security risks and considerations. Adding IoT devices previously located within the network to remote locations means businesses must protect new devices and endpoints from a variety of external security threats.

Every company has to explore unique technical infrastructure and security needs of the Edge computing infrastructure.

Here are the set of best practices that are emerging to improve IIoT network security with edge computing.

Identity definite Responsibility matrix for Edge Security

Enterprise digital organizations are split into two independent teams – Informational Technology (IT) and Operational Technology (OT). One of the challenging aspects of modern cybersecurity is the increasingly vague distinction between Informational Technology (IT) and Operational Technology (OT). Traditionally, all production equipment was the responsibility of OT, while network equipment was under the IT arena. Both organizations can manage the IIoT and edge computing devices now, as they reside at the edge but connect to centralized networks. IT-OT convergence is the best practice for an IIoT strategy. It is also critical for the security of the organization.

Companies must ensure alignment and collaboration between IT and OT teams on their divided responsibilities for securing the network. When possible, OT and IT teams should also understand each other’s roles and responsibilities so that they are capable of addressing any potential catastrophic events.

Perform a detailed Security Audit

Companies must first understand their vulnerabilities, to protect their operations from ransomware, endpoint manipulation, and other malicious security threats. The first critical step in improving network security is thoroughly mapping the IT infrastructure across the organization to identify all potential risks.

Corporate security teams should create an inventory of all assets, identify points of weakness, and build recovery plans for securing the network in the event of a security breach. Contracting an external security auditor will help organizations view their assets with an external perspective. Security audits should always involve both OT and IT teams. Similarly, effective recovery plans must establish the different roles of IT and OT in protecting company assets.

Select Security-Based Solutions

By design, IIoT and edge computing hardware face unique security challenges than traditional technologies. Security has not been a priority for purchasing or operational teams, as operational environments seldom pose a potential threat to company-wide networks.

Usually, production teams are focused on minimizing downtime and maximizing operational value. Hence they have a reliable set of priorities for hardware: simplicity, affordability, and autonomy, as required. However, the growth of IIoT devices and other increasingly smart operational equipment that is communicating with the network, make edge computing a new frontier in the cybersecurity war.

Many hardware manufacturers are finding unique ways to prioritize security without sacrificing operational simplicity. ‘Self-protecting’ edge computing and IIoT devices can autonomously identify data anomalies that might signal a cyber-attack or component malfunction. These devices reflect the increasing importance of edge security without introducing too much complexity for production teams. With self-protecting hardware, operational teams can ensure safety at the edge without sacrificing their ability to keep critical equipment up and running. Hence security is an important use case of Edge Machine Learning and Edge Artificial Intelligence.

Summary:

Edge Computing for IIoT: As companies explore the value of edge computing and increase the intelligence and efficiency of their industrial operations, they must remain aware of the potential security vulnerabilities. Developing a comprehensive cybersecurity plan requires full consideration of a company’s computing infrastructure, a clear awareness of security responsibilities, and invest in self-protecting hardware. Companies must first ensure they are adequately protected to maximize the value of IIoT and edge computing projects.

Footnotes:

Mobodexter, Inc based in Redmond, WA builds IOT Edge solutions for enterprises applications on Kubernetes & Dockers.

Register now at developers.paasmer.co and enjoy free Edge trial license that can run on Raspberry Pi.

Follow our weekly IoT Blogs: https://blogs.mobodexter.com

Join our IoT Hub of 3000+ subscribers to get these blogs in Email –  Subscribe.

Become our affiliate partner today.

Mobodexter | Mobodexter

Edge computing, the wide-ranging types of calculations and analyses at the extreme ends of an IT topology, has become a hot trend. According to TrendForce, a market research firm, the global market for edge computing is predicted to grow at CAGR of above 30% between now and 2022.

By 2025, the market size is expected to reach $3.24 billion. Also, edge computing is increasingly catching on in the auto industry. There are two distinct areas of application.

First one relates to all the in-vehicle computing operations, which are expanding rapidly with the rise of new generations of networked vehicles. Second, edge computing is making inroads into manufacturing processes, where growing concentrations of computing power moves as close to sensors as possible.

There are already a few critical projects in the connected-car space. BMW’s new car-as-a-sensor-service (Carasso) was developed in six months. It uses various services of AWS to achieve a comprehensive functionality. Using AWS, Carasso can adjust to very rapidly changing load requirements. In just 24 hours, they can go up or down two orders of magnitude.

Volkswagen has entered into a strategic partnership with Microsoft in the connected-car area. The new Volkswagen automotive cloud is expected to cover all the future digital services and mobility offerings from Volkswagen group. Their cloud is one sizeable industrial cloud, thereby creating the most significant digital ecosystem in the auto sector. In the future, all the services in VW vehicles and the group-wide, cloud-based carrier architecture One Digital Platform (ODP) are based on the Microsoft cloud platform Azure.

Automotive Edge Consortium

During 2018 summer, Ericsson established the Automotive Edge Computing Consortium (AECC). The cross-industry alliance, which includes AT&T, Denso, Intel, KDDI, NTT, Sumitomo Electric, and Toyota, wants to ensure that the new technologies and standards meet the future requirements of the value creation chain for networked cars.

The alliance assumes that networked cars will be sharing 10,000 times the current quantities of data by 2025. This data growth is a critical issue, as developments in the production machinery and robotics fields have shown. Edge computing has been expanding there with systems from a variety of manufacturers and with proprietary protocols. This variety creates a substantially increases complexity. Hence intermediate steps are needed to standardize data and processes.

Amazon is a very active player here. AWS recently announced four new services designed to simplify data capture in the IoT and edge computing fields. They are IoT SiteWise, IoT Events, IoT Things Graph, and IoT Greengrass Connector.

AWS services for Automated Cars

IoT SiteWise is a managed service for data capture and data cataloging. A preview version of SiteWise is now available and is currently being beta tested.

IoT Events supports responses to incidents generated by IOT sensors and applications. For example, if the temperature rises in a cold room, the change can be an indication that a door has been properly closed. A text is immediately sent to the responsible technician who can then look into the problem. The service is also only available as a preview version now.

IoT Things Graph can be used to create IoT apps on a visual basis with a drag-and-drop feature. Besides, there is a library with predefined cloud and device models. Amazon believes they can be used to develop applications with minimal coding.

The new IOT Greengrass Connector is available immediately. It helps app IoT developers connect external application securely to Greengrass devices.

Summary:

Autonomous cars equipped with Edge computing devices are seen as one large Edge computing entity that interacts with cloud computing for achieving various functionality. Hence cloud player like AWS is launching a few services to address the cloud side of this requirement.

Footnotes:

Mobodexter, Inc based in Redmond, WA builds IOT Edge solutions for enterprises applications on Kubernetes & Dockers.

Register now at developers.paasmer.co and enjoy free Edge trial license that can run on Raspberry Pi.

Follow our weekly IoT Blogs: https://blogs.mobodexter.com

Join our IoT Hub of 3000+ subscribers to get these blogs in Email –  Subscribe.

Become our affiliate partner today.