Category

Trend

Pushing AI to the Edge (Part Two): Edge AI in Practice and What’s Next

By Blog, LF Edge, Project EVE, Trend

Q&A with Jason Shepherd, LF Edge Governing Board member, Project EVE leader and VP of Ecosystem at ZEDEDA

Image for post

This content originally ran on the ZEDEDA Medium Blog – visit their website for more content like this.

In Part One of this two-part Q&A series we highlighted the new LF Edge taxonomy that publishes next week and some key considerations for edge AI deployments. In this installment our questions turn to emerging use cases and key trends for the future.

To discuss more on this budding space, we sat down with our Vice President of ecosystem development, Jason Shepherd, to get his thoughts on the potential for AI at the edge, key considerations for broad adoption, examples of edge AI in practice and some trends for the future.

What do you see as the most promising use cases for edge AI?

As highlighted in Part One, the reasons for deploying AI at the edge include balancing needs across the vectors of scalability, latency, bandwidth, autonomy, security and privacy. In a perfect world all processing would be centralized, however this breaks down in practice and the need for AI (and ML) at the edge will only continue to grow with the explosion of devices and data.

Hands down, computer vision is the killer app for edge AI today due to the bandwidth associated with streaming video. The ability to apply AI to “see” events in the physical world enables immense opportunity for innovation in areas such as object recognition, safety and security, quality control, predictive maintenance and compliance monitoring.

Considering retail — computer vision solutions will usher in a new wave of personalized services in brick and mortar stores that provide associates with real-time insights on current customers in addition to better informing longer term marketing decisions. Due to privacy concerns, the initial focus will be primarily around assessing shopper demographics (e.g., age, gender) and location but increasingly we’ll see personalized shopping experiences based on individual identity with proper opt-in (often triggered through customer loyalty programs). This includes a trend for new “experiential” shopping centers, for which customers expect to give up some privacy when they walk in the door in exchange for a better experience.

While Amazon Go stores have led the trend for autonomous shopping environments, the use of computer-vision enabled self-service kiosks for grab-and-go checkout is growing rapidly overall. Given the recent health concerns with COVID-19, providers are rapidly shifting to making these solutions contactless by leveraging gesture control, instead of requiring interaction with a keypad or touch screen.

Computer vision use cases will often leverage sensor fusion, for example with barcode scans or radio-frequency identification (RFID) technology providing additional context for decision making in retail inventory management and point of sale (POS) systems. A camera can tell the difference between a T-shirt and a TV, but not the difference between large and medium sizes of the same shirt design. Still, perhaps eventually an AI model will be able to tell you if you have bad taste in clothing!

Another key vertical that will benefit from computer vision at the edge is healthcare. At ZEDEDA, we’ve worked with a global provider that leverages AI models in their medical imaging machines, located within hospitals and provided as a managed service. In this instance, the service provider doesn’t own the network on which their machines are deployed so they need a zero-trust security model in addition to the right tools to orchestrate their hardware and software updates.

Another example where bandwidth drives a need for deploying AI at the IoT Edge is vibration analysis as part of a use case like predictive maintenance. Here sampling rates of at least 1KHz are common, and can increase to 8–10KHz and beyond because these higher resolutions improve visibility into impending machine failures. This represents a significant amount of continuously streaming data that is cost-prohibitive to send directly to a centralized data center for analysis. Instead, inferencing models will be commonly deployed on compute hardware proximal to machines to analyze the vibration data in real time and only backhauling events highlighting an impending failure.

Analysis for predictive maintenance will also commonly leverage sensor fusion by combining this vibration data with measurements for temperature and power (voltage and current). Computer vision is also increasingly being used for this use case, for example the subtle wobble of a spinning motor shaft can be detected with a sufficient camera resolution, plus heat can be measured with thermal imaging sensors. Meanwhile, last I checked voltage and current can’t be checked with a camera!

An example of edge AI served up by the Service Provider Edge is for cellular vehicle-to-everything (C-V2X) use cases. While latency-critical workloads such as controlling steering and braking will always be run inside of a vehicle, service providers will leverage AI models deployed on compute proximal to small cells in a 5G network within public infrastructure to serve up infotainment, Augmented Reality for vehicle heads-up displays and coordinating traffic. For the latter, these AI models can warn two cars that they are approaching a potentially dangerous situation at an intersection and even alert nearby pedestrians via their smartphones. As we continue to collaborate on foundational frameworks that support interoperability it will open up possibilities to leverage more and more sensor fusion that bridges intelligence across different edge nodes to help drive even more informed decisions.

We’re also turning to processing at the edge to minimize data movement and preserve privacy. When AI inferencing models are shrunk for use in constrained connected products or healthcare wearables, we can train local inferencing models to redact PII before data is sent to centralized locations for deeper analysis.

Who are the different stakeholders involved in accelerating adoption of edge AI?

Edge AI requires the efforts of a number of different industry players to come together. We need hardware OEMs and silicon providers for processing; cloud scalers to provide tools and datasets; telcos to manage the connectivity piece; software vendors to help productize frameworks and AI models; domain expert system integrators to develop industry-specific models, and security providers to ensure the process is secure.

In addition to having the right stakeholders it’s about building an ecosystem based on common, open frameworks for interoperability with investment focused on the value add on top. Today there is a plethora of choices for platforms and AI tools sets which is confusing, but it’s more of the state of the market than necessity. A key point of efforts like LF Edge is to work in the open source community to build more open, interoperable, consistent and trusted infrastructure and application frameworks so developers and end users can focus on surrounding value add. Throughout the history of technology open interoperability has always won out over proprietary strategies when it comes to scale.

In the long run, the most successful digital transformation efforts will be led by organizations that have the best domain knowledge, algorithms, applications and services, not those that reinvent foundational plumbing. This is why open source software has become such a critical enabler across enterprises of all sizes — facilitating the creation of de-facto standards and minimizing “undifferentiated heavy lifting” through a shared technology investment. It also drives interoperability which is key for realizing maximum business potential in the long term through interconnecting ecosystems… but that’s another blog for the near future!

How do people differentiate with AI in the long term?

Over time, AI software frameworks will become more standardized as part of foundational infrastructure, and the algorithms, domain knowledge and services on top will be where developers continue to meaningfully differentiate. We’ll see AI models for common tasks — for example assessing the demographics of people in a room, detecting license plate numbers, recognizing common objects like people, trees, bicycles and water bottles — become commodities over time. Meanwhile, programming to specific industry contexts (e.g. a specific part geometry for manufacturing quality control) will be where value is continually added. Domain knowledge will always be one of the most important aspects of any provider’s offering.

What are some additional prerequisites for making edge AI viable at scale?

In addition to having the right ecosystem including domain experts that can pull solutions together, a key factor for edge AI success is having a consistent delivery or orchestration mechanism for both compute and AI tools. The reality is that to date many edge AI solutions have been lab experiments or limited field trials, not yet deployed and tested at scale. PoC, party of one, your table is ready!

Meanwhile, as organizations start to scale their solutions in the field they quickly realize the challenges. From our experience at ZEDEDA, we consistently see that manual deployment of edge computing using brute-force scripting and command-line interface (CLI) interaction becomes cost-prohibitive for customers at around 50 distributed nodes. In order to scale, enterprises need to build on an orchestration solution that takes into account the unique needs of the distributed IoT edge in terms of diversity, resource constraints and security, and helps admins, developers and data scientists alike keep tabs on their deployments in the field. This includes having visibility into any potential issues that could lead to inaccurate analyses or total failure. Further, it’s important that this foundation is based on an open model to maximize potential in the long run.

Where is edge AI headed?

To date, much of the exploration involving AI at the edge has been focused on inferencing models — deployed after these algorithms have been trained with the scalable compute of the cloud. (P.S. for those of you who enjoy a good sports reference, think of training vs. inference as analogous to coaching vs. playing).

Meanwhile, we’re starting to see training and even federated learning selectively moving to the Service Provider and User Edges. Federated learning is an evolving space that seeks to balance the benefits of decentralization for reasons of privacy, autonomy, data sovereignty and bandwidth savings, while centralizing results from distributed data zones to eliminate regional bias.

The industry is also increasingly developing purpose-built silicon that can increase efficiencies amid power and thermal constraints in small devices and even support either training or inference and this corresponds with the shift towards pushing more and more AI workloads onto edge devices. Because of this, it’s important to leverage device and application orchestration tools that are completely agnostic to silicon, compared to offers from silicon makers that have a vested interest in locking you into their ecosystem.

Finally, we’ll see the lower boundary for edge AI increasingly extend into the Constrained Device Edge with the rise of “Tiny ML” — the practice of deploying small inferencing models optimized for highly constrained, microcontroller-based devices. An example of this is the “Hey Alexa” of an Amazon Echo that is recognized locally and subsequently opens the pipe to the cloud-based servers for a session. These Tiny ML algorithms will increasingly be used for localized analysis of simple voice and gesture commands, common sounds such as a gunshot or a baby crying, assessing location and orientation, environmental conditions, vital signs, and so forth.

To manage all of this complexity at scale, we’ll lean heavily on industry standardization, which will help us focus on value on top of common building blocks. Open source AI interoperability projects, such as ONNX, show great promise in helping the industry coalesce around a format so that others can focus on developing and moving models across frameworks and from cloud to edge. The Linux Foundation’s Trust over IP effort and emerging Project Alvarium will also help ease the process of transporting trusted data from devices to applications. This notion of pervasive data trust will lead to what I call the “Holy Grail of Digital” — selling and/or sharing data resources and services to/with people you don’t even know. Now this is scale!

In Closing

As the edge AI space develops, it’s important to avoid being locked into a particular tool set, instead opting to build a future-proofed infrastructure that accommodates a rapidly changing technology landscape and that can scale as you interconnect your business with other ecosystems. Here at ZEDEDA, our mission is to provide enterprises with an optimal solution for deploying workloads at the IoT Edge where traditional data center solutions aren’t applicable, and we’re doing it based on an open, vendor-neutral model that provides freedom of choice for hardware, AI framework, apps and clouds. We’re even integrating with major cloud platforms such as Microsoft Azure to augment their data services.

Reach out if you’re interested in learning more about how ZEDEDA’s orchestration solution can help you deploy AI at the IoT Edge today while keeping your options open for the future. We also welcome you to join us in contributing to Project EVE within LF Edge which is the open source foundation for our commercial cloud offering. The goal of the EVE community is to build the “Android of the IoT Edge” that can serve as a universal abstraction layer for IoT Edge computing — the only foundation you need to securely deploy any workload on distributed compute resources. To this end, a key next step for Project EVE is to extend Kubernetes to the IoT Edge, while taking into account the unique needs of compute resources deployed outside of secure data centers.

The success of AI overall — and especially edge AI — will require our concerted collaboration and alignment to move the industry forward while protecting us from potential misuse along the way. The future of technology is about open collaboration on undifferentiated plumbing so we can focus on value and build increasingly interconnected ecosystems that drive new outcomes and revenue streams. As one political figure famously said — “it takes a village!”

If you have questions or would like to chat with leaders in the project, join us on the LF Edge Slack  (#eve or #eve-help) or subscribe to the email list. You can check out the documentation here.

Exploration and Practices of Edge Computing: Cloud Managing Containerized Devices

By Blog, EdgeX Foundry, Industry Article, Trend

Written by Gavin Lu, LF Edge member, EdgeX Foundry China Project Lead and R&D Director in the VMware Office of the CTO

As an industry leader with vast experience and knowledge, Gavin has been writing a series of articles focused on edge computing. These articles are posted on his personal blog and are posted here with his permission. To read more content from Gavin, visit his website.

Introduction

The previous article introduced the cloud management virtualization device solution. This article will describe the Nebula project, a unified management of containerized devices and edge applications and data analysis cloud services.

Nebula Architecture

Project Nebula is designed based on the following key ideas:

  • Agnostic to device CPU architecture, supporting both x86 and ARM;
  • Agnostic to edge application frameworks, supporting EdgeX Foundry and other frameworks that can be packaged and run;
  • Agnostic to data analytics services, supporting on-premise and cloud deployment;
  • Support small to large scale deployment;
  • Support end-to-end multi-tenant operation model from device to cloud.
EdgeX Foundry Architecture

Nebula supports EdgeX Foundry framework, and we already published a live test bed at https://18.189.42.126/. Those who are interested in Nebula could contact yixingj@vmware.com to register for a trial, installation and user guides with detailed information.

Nebula Demo

Installation

Nebula is designed in containerized micro-service architecture, and is installed by default in OVA format. Similar to Pallas architecture introduced in the previous article, although Nebula package is encapsulated in OVA, it does not depend on any specific virtualization infrastructure or cloud platform to be installed. Technically, it could completely be converted to other formats, or install on any cloud platform that supports OVA format.

The basic resource requirement of Nebula is:

  • CPU: 2 virtual CPU cores
  • Memory: 8GB
  • Storage: 150GB

Its installation process is similar to other normal OVA, and users can log in as the administrator after completion.

Nebula Service Console

Vendor Portal

After the installation is complete, users can log in to the vendor portal as an administrator according to the prompt address in VM console as above and perform user management.

Nebula Management Portal

In Nebula, edge application services are defined as following: A Service can contain multiple Versions, and a Version contains multiple Service Components.

Edge Service Hierarchy

For each service created, it is necessary to determine parameters and resource requirement such as version, CPU platform, memory, storage, network, etc., to facilitate verification in full life cycle management.

Vendors can upload a set of EdgeX Foundry applications packaged in container images, and define categories, dependencies between containers, resource parameters, startup order, and parameters of connected data analysis cloud services.

After the release, users can see and deploy these edge services.

Device Registration

Before users actually deploy EdgeX Foundry applications, they must first register the device they would use into their Nebula accounts.

Users need to download Nebula agent program nebulacli.tar by themselves and run it on the device to complete the registration. This registration step could be manual, or it can be automated in batch operations for OEM.

./install.sh init -u user-acccount -p user-account-password -n user-device-name

User Portal

After completing the device registration, users can install and manage EdgeX Foundry or other edge applications released in advance on Nebula service by vendors. Users can find proper applications in the catalog.

After selection, users can further specify parameter settings of the deployment in the drag-and-drop wizard, which maps to parameter values defined by the vendor before.

After all parameters are set, the actual deployment can be carried out, either in batch or multiple times to multiple devices. After deploying EdgeX Foundry applications, users can monitor device resources and application run time status in real time.

Nebula provides complete Restful API documentation, with which users can automate operations to deploy EdgeX Foundry applications in a large scale.

Next

From the second article to this article, I introduced the basic method of building and managing virtualized devices and containerized devices from the cloud. But I did not answer the question of how to deal with single-point device failure. Compared with the traditionally inflexible and inefficient full redundancy or external NAS solution, the next article will introduce device clusters on hyper-convergence architecture.

Pushing AI to the Edge (Part One): Key Considerations for AI at the Edge

By Blog, LF Edge, Project EVE, State of the Edge, Trend

Q&A with Jason Shepherd, LF Edge Governing Board member and VP of Ecosystem at ZEDEDA

This content originally ran on the ZEDEDA Medium Blog – visit their website for more content like this.

This two-part blog provides more insights into what’s becoming a hot topic in the AI market — the edge. To discuss more on this budding space, we sat down with our Vice President of ecosystem development, Jason Shepherd, to get his thoughts on the potential for AI at the edge, key considerations for broad adoption, examples of edge AI in practice and some trends for the future.


Chart defining the categories within the edge, as defined by LF Edge

Image courtesy of LF Edge

Breaking Down the Edge Continuum

By Blog, State of the Edge, Trend, Use Cases

Written by Kurt Rinehart, Director of Data Science at Section. This blog originally ran on the Section website. For more content like this, please click here.

There are many definitions of “the edge” out there. Sometimes it can seem as if everyone has their own version.

LF Edge, an umbrella organization that brings together industry leaders to build “an open source framework for the edge,” has a number of edge projects under its remit, each of which seeks to unify the industry around coalescing principles and thereby accelerate open source edge computing developments. Part of its remit is to define what the edge is, an invaluable resource for the edge community to coalesce around.

Latest LF Edge White Paper: Sharpening the Edge

In 2018, State of the Edge (which recently became an official project of LF Edge) put out its inaugural report, defining the edge using four criteria:

  • “The edge is a location not a thing;
  • There are lots of edges, but the edge we care about today is the edge of the last mile network;
  • This edge has two sides: an infrastructure edge and a device edge;
  • Compute will exist on both sides, working in coordination with the centralized cloud.”

Since that inaugural report, much has evolved within the edge ecosystem. The latest white paper from LF Edge, Sharpening the Edge: Overview of the LF Edge Taxonomy and Framework, expands on these definitions and moves on from simply defining two sides (the infrastructure and the device edge) to use the concept of an edge continuum.

The Edge Continuum

The concept of the edge continuum describes the distribution of resources and software stacks between centralized data centers and deployed nodes in the field as “a path, on both the service provider and user sides of the last mile network.”

In almost the same breath, LF Edge also describes edge computing as essentially “distributed cloud computing, comprising multiple application components interconnected by a network.”

We typically think of “the edge” or “the edges” in terms of the physical devices or infrastructure where application elements run. However, the idea of a path between the centralized cloud (also referred to as “the cloud edge” or “Internet edge”) and the device edge instead allows for the conceptualization of multiple steps along the way.

The latest white paper concentrates on two main edge categories within the edge continuum: the Service Provider Edge and the User Edge (each of which is broken down into further subcategories).

edge continuum diagram
Image source: LF Edge

The Service Provider Edge and the User Edge

LF Edge positions devices at one extreme of the edge continuum and the cloud at the other.

Next along the line of the continuum after the cloud, also described as “the first main edge tier”, is the Service Provider (SP) Edge. Similarly to the public cloud, the infrastructure that runs at the SP Edge (compute, storage and networking) is usually consumed as a service. In addition to the public cloud, there are also cellular-based solutions at the SP Edge, which are typically more secure and private than the public cloud, as a result of the differences between the Internet and cellular systems. The SP Edge leverages substantial investments by Communications Service Providers (CSPs) into the network edge, including hundreds of thousands of servers at Points of Presence (PoPs). Infrastructure at this edge tier is largely more standardized than compute at the User Edge.

The second top-level edge tier is the User Edge, which is on the other side of the last mile network. It represents a wider mix of resources in comparison to the SP Edge, and “as a general rule, the closer the edge compute resources get to the physical world, the more constrained and specialized they become.” In comparison to the SP Edge and the cloud where resources are owned by these entities and shared across multiple users, resources at the User Edge tend to be customer-owned and operated.

Moving from the Cloud to the Edge

What do we mean when we talk about moving from the cloud to the edge? Each of the stages along the edge continuum take you progressively closer to the end user. You have high latency and more compute in the centralized cloud versus low latency and less compute as you get closer to the User Edge. When we talk about moving from the cloud to the edge, it means we want to leverage the whole stack and not solely focus on the centralized cloud.

Let’s look at the most obvious use case: content delivery networks (CDNs). In the 1990s, Akamai created content delivery networks to allow localized websites to serve a global audience. A website based in New York could leverage Akamai’s distributed network of proxy servers and data centers around the world to be able to store their static assets globally, including HTML, CSS, JavaScript, video, and images. By caching these in Akamai’s distributed global points of presence (PoP), the website’s end users worldwide were guaranteed high availability and consistent performance.

These days, CDNs are considered to be only one layer in a highly complex Internet ecosystem. Content owners such as media companies and e-commerce vendors continue to pay CDN operators to deliver their content to end users. In turn, a CDN pays ISPs, carriers, and network operators for hosting its servers in their data centers. That’s the Service Provider Edge we’re talking about.

An edge compute platform is still a geographically distributed network, but instead of simply providing proxy servers and data centers, an edge compute platform also offers compute. How do we define this? Compute can be defined as many things, but essentially, it boils down to the ability to run workloads wherever you need to run them. Compute still gives you high availability and performance, but it also allows for the capability to run packaged and custom workloads positioned relatively spatially to users.

An edge compute platform leverages all available compute between the cloud provider and the end user, together with DevOps practices, to deliver traditional CDN and custom workloads.

Applying Lessons from the Cloud to the Edge

We can take the lessons we’ve learned in the cloud and apply them to the edge. These include:

  • Flexibility – At Section, we describe this as wanting to be able to run “any workload, anywhere”, including packaged and customized workloads;
  • Taking a multi-provider approach to deployments – This offers the opportunity to create a higher layer of abstraction. Infrastructure as Code (IaC) is the process of managing and provisioning computer data centers through machine-readable definition files as opposed to physical hardware configuration or interactive configuration tools. At Section, we have 6-7 different providers, from cloud providers to boutique providers to bare metal providers.
  • Applying DevOps practices – In order to provide the capabilities that the cloud has at the infrastructure edge, we need to enable developers to get insight and to run things at the edge at speed, just as they did in the cloud. This is DevOps. It’s important to be able to apply DevOps practices here since, “if you build it, you own it”. You want to make things open, customizable, and API-driven with integrations, so that developers can leverage and build on top of them.
  • Leveraging containerized workloads – Deploying containers at the edge involves multiple challenges, particularly around connectivity, distribution and synchronization, but it can be done, and in doing, allows you to leverage this architecture to deploy your own logic, not just pre-packaged ones. Containerization also offers:
    • Security
    • Standardization
    • Isolation; and
    • A lightweight footprint.
  • Insights and Visibility – We need to give developers deep, robust insight into what’s happening at the edge, just as we do in the cloud. The three pillars of observability are logs, metrics and tracing. An ELK stack can provide this, giving developers the invaluable ability to understand what is happening when things inevitably go wrong.

Edge Computing Use Cases in the Wild

There are many examples of use cases already operating at the Edge. A few of the many interesting ones out there include:

  • Facebook Live – When you see a live stream in your feed and click on it, you are requesting the manifest. If the manifest isn’t already on your local PoP, the request travels to the data center to get the manifest, and then fetches the media files in 1 second clips. ML algorithms operate on the 1 second clips to optimize them in real time to deliver the best, fastest experience for users.
  • Cloudflare Workers – These are Service Worker API implementations for the Cloudflare platform. They deploy a server-side approach to running JavaSCript workloads on Cloudflare’s global network.
  • Chick-fil-A – A surprising one. Chick-fil-A has been pushing into the device edge over the last couple of years. Each of their 20,000 stores has a Kubernetes cluster that runs there. The goal: “low latency, Internet-independent applications that can reliably run our business”, in addition to high availability for these applications, a platform that enables rapid innovation, and the ability to horizontally scale.

We’re Not Throwing Away the Cloud

One last thing to make clear: we’re not talking about throwing away the cloud. The cloud is going nowhere. We will be working alongside it, using it. What we’re talking about is moving the boundary of our applications out of the cloud closer to the end user, into the compute that is available there. And, as we’ve seen, we don’t need to throw away the lessons we’ve learned in the cloud; we can still use the tools that we’re used to, plus gain all the advantages that the edge continuum has to offer.

You can download the LF Edge taxonomy white paper here. You can also watch the LF Edge Taxonomy Webinar, which shares insight from the white paper, on our Youtube Channel. Click here to watch it now.  

Enabling a More Sustainable Energy Model with Edge Computing

By Blog, Open Glossary of Edge Computing, Trend

Written by Molly Wojcik, Chair of the State of the Edge Landscape Working Group and Marketing Director at Section

This blog previously ran on the Section website. For more content like this, click here.

The tech sector has been under mounting scrutiny over the last decade by environmental groups, businesses, and customers for its increasing energy consumption levels. Attention has recently been turning to the ways in which edge computing can help build more sustainable solutions.

The Impact of Coronavirus Lockdowns on Internet Usage

While worldwide energy consumption has significantly dropped as a result of global lockdown restrictions, Internet usage has seen a huge spike.

NETSCOUT, a provider of network and application performance monitoring tools, saw a 25-35% increase in worldwide Internet traffic patterns in mid-March, the time when the shift to remote work and online learning happened for much of the world’s population. That number has stayed pretty consistent. There has been particularly increased use of bandwidth-intensive applications like video streaming, Zoom and online gaming. The spike in Internet use has implications for the sector’s sustainability targets and makes the urgency of a call for change even louder.

Before we look specifically at sustainability in the edge computing field, let’s pull back to take a look at the wider sector.

The Energy Challenge

The information and communication technology (ICT) ecosystem accounts for over 2% of global emissions, putting it on a par with the aviation industry’s carbon footprint, at least according to a 2018 study. Did you know that data centers use an estimated 200 terawatt (Twh) hours each year? This represents around 1% of global electricity demand and data centers contribute around 0.3% to overall carbon emissions.

Where these figures could lie in the future is harder to predict. Widely cited forecasts say the ICT sector and data centers will take a larger slice of overall electricity demand. Some experts predict this could rise as high as 8% by 2030, while the most aggressive models predict that electricity usage by ICT could surpass 20% of the global total by the time a child born today hits his or her teens, with data centers using over one-third of that.

Keeping Future Energy Demand in Check

However, improvements in energy efficiency in the data center field are already making a difference. While the amount of computing performed in data centers more than quintupled between 2010 and 2018, the amount of energy consumed by data centers worldwide only rose by 6% across the same period.

The ICT sector has been hard at work to keep future energy demand in check in various ways. These include:

Streamlining Computer Processes

The shift away from legacy enterprise data centers operated by traditional enterprises such as banks, retailers, and insurance companies, to newer commercially operated cloud data centers has been making a big difference to overall energy consumption.

In a recent company blog, Urs Hölzle, Google’s senior VP technical infrastructure, pointed to a study in Science showing the gains in energy efficiency. Hölzle wrote, “a Google data center is twice as energy efficient as a typical enterprise data center. And compared with five years ago, we now deliver around seven times as much computing power with the same amount of electrical power.”

The study shows how an overall shift to hyperscale data centers has helped reduce the amount of traditional enterprise data center capacity and subsequently reduced overall energy consumption. One of the authors is Jonathan Koomey, a former Lawrence Berkeley National Laboratory scientist, who has been studying the subject of data center energy usage for over two decades. Koomey says that forecasts that focus on data growth projections alone ignore the energy efficiency gains the sector has been making.

Facebook has also been exploring ways to maximise efficiency over the last decade. The social media giant started the Open Compute Project in 2011 to share hardware and software solutions aimed at making computing more energy-efficient.

One recent novel idea is that data centers could act as energy suppliers and sell excess electricity to the grid instead of keeping it as an insurance policy. Kevin Hagen, Iron Mountain’s VP of environment, social and governance strategy, describes backup generators and UPS batteries as “useless capital.”

Instead, he asks, “What would it look like if all that money was actually invested in energy systems that we got to use when we were trying to arbitrage energy during the day and night on a regular basis? What if we share control with the utility, so they can see there’s an asset and use it back and forth?”

New Ways to Cool Data Centers

According to Global Market Insights, cooling systems represent on average 40% of entire data center energy consumption. Data center design specialists have been looking into different ways to approach reducing energy needs specifically for cooling, which have been largely approached in the same way for the last three decades. These include locating data centers in cooler areas, using AI to regulate the data center’s cooling systems to match the weather, warm-water cooling, immersion cooling, and rear door cooling systems.

It’s an important problem to be focused on since data centers worldwide contribute to industry’s consumption of 45% of all available clean water.

A Focus on Sustainables

Greenpeace has been putting data centers under the spotlight for over a decade, calling for them and other digital infrastructure to become 100% renewably powered. There has been great progress since 2010 when Greenpeace started its ClickClean campaign when IT companies were negligible contributors to renewable-power purchase agreements; today Google is the world’s largest corporate purchaser of renewable energy.

Apple was at the top of Greenpeace’s list of the top greenest tech companies last year. It already boasts 100% renewable energy to power all its global data centers, and is currently focused on cleaning up its supply chain. Greenpeace has recently called out AWS, however, for a lack of renewables in Virginia’s “data center alley”, which it says is powered by “dirty energy.”

Corporate data centers have traditionally been part of the “dirty energy” problem, but are beginning to participate in seeking out renewable energy. Some of the initiatives in this space include DigitalRealty’s announcement on using wind energy in its 13 Dallas, TX data centers and IronMountain’s Green Power Pass specifically for enterprise customers to be able to participate in renewable energy purchase for its growing number of data centers.

The Role of Edge Computing in Working Towards More Efficient Solutions

Edge computing is also increasingly being looked to as an area that can help work towards more efficient solutions. These include:

Energy Resource Efficiencies

technical paper published on Arxiv.org earlier this month lifted the hood on Autoscale, Facebook’s energy-sensitive load balancer. Autoscale reduces the number of servers that need to be on during low-traffic hours and specifically focuses on AI, which can run on smartphones in abundance and lead to decreased battery life through energy drain and performance issues.

The Autoscale technology leverages AI to enable energy-efficient inference on smartphones and other edge devices. The intention is for Autoscale to automate deployment decisions and decide whether AI should run on-device, in the cloud, or on a private cloud. Autoscale could result in both cost and efficiency savings.

Other energy resource efficiency programs are also underway in the edge computing space.

Reducing High Bandwidth Energy Consumption

Another area edge can assist in is to reduce the amount of data traversing the network. This is especially important for high-bandwidth applications like YouTube and Netflix, which have skyrocketed in recent years, and recent months in particular. This is partly due to the fact each stream is composed of a large file, but also due to how video-on-demand content is distributed in a one-to-one model. High bandwidth consumption is linked to high energy usage and high carbon emissions since it uses the network more heavily and demands greater power.

Edge computing could help optimize energy usage by reducing the amount of data traversing the network. By running applications at the user edge, data can be stored and processed close to the end user and their devices instead of relying on centralized data centers that are often hundreds of miles away. This will lead to lower latency for the end user and could lead to a significant reduction in energy consumption.

Smart Grids and Monitoring Can Lead to Better Management of Energy Consumption

Edge computing can also play a key role in being an enabler of solutions that help enterprises better monitor and manage their energy consumption. Edge compute already supports many smart grid applications, such as grid optimization and demand management. Allowing enterprises to track and monitor energy usage in real-time and visualize it through dashboards, enterprises can better manage their energy usage and put preventative measures in place to limit it where possible. This kind of real-time assessment of supply and demand can be particularly useful for managing limited renewable energy resources, such as wind and solar power.

Examples in the Wild

Schneider Electric

Schneider Electric, a specialist in energy management and automation, is beginning to see clients seek more efficient infrastructure and edge computing solutions. This includes helping organizations to digitize their manufacturing processes by using data to produce real-time feedback and alerts to gain efficiencies across the supply chain, including the reduction of waste and carbon emissions. Natalya Makarochkina, Senior VP, International, Secure Power at Schneider Electric, says edge computing is allowing this to happen in a scalable and sustainable way.

Makarochkina highlights two Schneider customers that have used edge computing to improve sustainability and make efficiency savings. The first is a specialty grocer that used edge computing to upgrade IT infrastructure and reduce its physical footprint and consumption requirements, which led to a 35% reduction in engineering cost and a 50% increase in deployment speed for the end user. The second is a co-working space provider that used Schneider’s edge infrastructure options to launch an on-site data center that offers IT infrastructure on demand, offering tenants greater operational efficiencies.

“Edge computing and hybrid cloud play a pivotal role in sustainability because they are designed specifically to put applications and data closer to devices and their users. This helps organisations to better react to changes in consumer demands and improve processes to create more sustainable products.”– Natalya Makarochkina, Senior VP, International, Secure Power at Schneider Electric

Section

At Section, we’re working to provide more accessible options for developers to build edge strategies that deliver on sustainability targets. Our patent-pending Adaptive Edge Engine technology allows developers to piece together bespoke edge networks that are optimized across various attributes, with carbon-neutral underlying infrastructure included in the portfolio of options. Other attributes include cost, performance, and security/regulatory requirements.

Similarly to Facebook’s Autoscale, Section’s Adaptive Edge Engine tunes the required compute resources at the edge in response to, not only the application’s required compute attributes (e.g. PCI compliance or carbon neutrality), but also continuously in response to traffic served and performance. Our ML forecasting engine sits behind the workload placement and decision engine which continuously places workload and scales the underlying infrastructure (up and down) so we use the least amount of compute while providing the maximum availability and performance.

Summary

Not many would argue that technology’s immersion in our day-to-day lives is only expected to expand. Technology providers must work together to create sustainable solutions to meet these growing demands. While not an end-all solution, edge computing has the potential to play a key role in helping to control the negative impacts that accompany rising energy demands.

Section is a LF Edge Member. To learn more about LF Edge or any of our members, visit https://www.lfedge.org/members/.

How Open Source is Driving 5G, Edge, AI and IoT

By Blog, Training, Trend

The 5G transition is well underway, with the technology rolled out on every continent, and adoption growing daily. This is leading to advances in other technologies – most especially edge computing, artificial intelligence and the Internet of Things. Many don’t realize that open source software is at the heart of the 5G revolution, making it possible in the first place and helping to speed implementation thanks to shared R&D efforts and greater interoperability than prior wireless standards. 

Considering the accelerating rate of change in the networking and telecommunications industry, it can be difficult to stay up to speed on these and the other latest technologies. Managers and their technical partners will be the ones to build the next great innovations based on the capabilities of 5G – but in order to do so, they require a fundamental understanding of the market pressures and a basic understanding of the technologies driving this shift – technologies like edge computing, IoT and AI.

That’s why The Linux Foundation offers two online training courses exploring these topics free of charge. Business Considerations for 5G, IoT, and AI is designed to help you discern between the hype and real opportunities of 5G technologies. Open Source and the 5G Transition explains the open source infrastructure powering the future and how to leverage it for business benefit. 

These courses are only two hours long, and no technical expertise is required. They are designed for anyone from business professionals to engineers who want to improve their understanding of these technologies and the changes they bring. Register for free today and increase your knowledge!