LEADERSHIP VOICES
Sponsored by
Sponsored Content:
Not so long ago, cloud computing changed the world as we knew it. It drove a rush to move workloads, applications and data into the cloud to take advantage of cost and management efficiencies. But a funny thing happened on the way to nirvana in the cloud, says Steve Orrin, Federal Chief Technology Officer at Intel: “A lot of organizations learned quickly that you don’t get anything for free.” It turns out, he says, “There’s a cost to doing things in the cloud, and if you don’t build cloud-native applications, you’re not really getting all the benefits you thought you’d get.”
This sentiment is especially true in the federal sector, where “lift-and-shift” cloud efforts — a common strategy for moving workloads from one environment to another — helped migrate data and applications quickly and easily, but didn’t necessarily take into account the steps they had to take to prepare for successful cloud adoption.
Orrin, who has spent several years in his role at Intel, knows a thing or two about the tools needed to collect, aggregate, analyze and respond to data before it is transmitted to a cloud environment. With a host of offerings designed specifically for government agencies and contractors as they embark on their cloud migration journey, Intel’s edge-to-cloud offerings are designed to satisfy specific government requirements. As a result, technologists like Orrin have a front-row seat on the trends and demand signals shaping the future of federal cloud services.
Government at the Edge
These trends, Orrin explains, are moving in two directions: First, there is the general drive to converge data centers and build and deploy cloud-native applications; and second, is the emergence of the Internet of Things, with a proliferation of sensors and endpoints that generate massive data loads and demand instant local processing to make that data useful.
“That data is being generated at the edge by devices and sensors,” Orrin says. “There’s a massive expense of transferring all of that data into the cloud to do anything useful with it. What we are now seeing is the next logical extension: Looking at mission applications, enterprise applications with the aim to drive intelligence and value to the end user in real time.”
To do that, agencies will need to ensure some processing occurs at the edge, he notes.
To derive the full value of artificial intelligence and machine learning, Orrin says intelligent edge computing is critical — especially where multitudes of data require advanced or accelerated processing capabilities. “If you want to do natural language processing for translation, or video or object recognition, you quickly see the need for doing something closer to the edge,” he explains.
“Data is being generated at the edge by devices and sensors. There’s a massive expense of transferring all of that data into the cloud to do anything useful with it. What we are now seeing is the next logical extension: Looking at mission applications, enterprise applications with the aim to drive intelligence and value to the end user in real time.”
STEVE ORRIN | Federal Chief Technology Officer, Intel
An Edge-to-Cloud Model to Enhance the User Experience
Users don’t want to wait for data to travel halfway across the country or around the world and back. Nor is that always possible, especially in national security or other remote applications where network connectivity may be unreliable and where mission effectiveness cannot be dependent on connectivity. Such systems must be able to operate when networks are overwhelmed, degraded or disconnected.
But adding cloud-like processing at the edge does not mean abandoning the cloud itself.
“It means we have to do the right computing and the right processing at the right place,” Orrin says. “Video recognition and object recognition is a really good use case to understand why this is: If you’re driving in a Tesla, you want to be able to recognize that car at the corner right now, not when the network is ready to respond. You can’t wait for the cloud to communicate back and say, ‘Oh, yes, that’s a car, don’t turn.’”
At the same time, users depend on access to the cloud to do their jobs effectively. For instance, a Transportation Security Services agent, a reconnaissance drone in the sky or a soldier at an observation post in a combat zone all require access to locally processed data in real time as well as cloud-fed data from other sources to gain valuable insight and make informed decisions quickly and effectively.
“Also, you want to combine data from across multiple mobile platforms, doing both sensing and analysis, and then combine those to do things like predictive traffic patterns, to recognize inputs and use them to make better decisions,” Orrin adds. “That’s where the cloud is really powerful, being able to go across multiple sensors, and do the Big Data large-scale analytics. Now you can get insights you don’t need in real time, as well as the information you need immediately. That’s really the power of the edge and the cloud working together.”
“You want to combine data from across multiple mobile platforms, doing both sensing and analysis, and then combine those to do things like predictive traffic patterns, to recognize inputs and use them to make better decisions.”
STEVE ORRIN | Federal Chief Technology Officer, Intel
Edge-to-Cloud as a Public Sector Strategy
The Internet of Things is really just a proliferation of edge devices, which vary more in terms of where they are than how they operate. One tangible example of the IoT revolution at work can be found inside the Department of Defense. To arm members of the Air Force with improved situational awareness and the ability to make game-time decisions rapidly and accurately, DoD adopted the Joint All-Domain Command and Control, a system that aims to support operations amid a variety of conflicts and often-dangerous or sophisticated environments. The goal of JADC2 is to ensure cars, aircraft, munitions, satellites, ships, submarines, tanks and people are at the right place at the right time.
Here’s how it works: Data needed locally can be processed locally, while unprocessed data can flow directly to wherever it is needed. True situational awareness is achieved by combining a dozen, 100 or even 1,000 data feeds into one cohesive whole.
“That’s the centralized view,” Orrin says, equating that part of the process to the cloud. “But in the cockpit of that airplane, you’re really operating a mobile edge computing platform, coordinating with back-end systems that provide the situational awareness you need to operate.”
Thanks to these technological advancements, an individual soldier can now carry a micro rack of servers in a pack or in the back of a vehicle that are able to do heavy processing, “like a micro cloud,” Orrin explains.
It’s not just the military that has leveraged these types of edge-to-cloud systems to empower heightened situational awareness and faster response times. In fact, the Federal Emergency Management Agency has similar needs.
“[First responders] get to a disaster zone and within a couple of minutes they’re operational – with all the capabilities they need for search and rescue, disaster relief, and other coordinated activities,” Orrin says. “That’s thanks to a micro cloud, with all of their key applications ready to go on a hyper-converged platform.”
This type of miniature clouds is a far cry from large commercial cloud offerings.
“On the edge, it’s not about infinite scale or elasticity, but rather the ability to dynamically spin up applications on demand, to build once, deploy everywhere, and to be able to use API-driven data models to deploy capabilities more quickly,” Orrin explained. “There’s also the ability to rapidly deploy new capabilities with existing data sets and access data more quickly.”
“[First responders] get to a disaster zone and within a couple of minutes they’re operational – with all the capabilities they need for search and rescue, disaster relief, and other coordinated activities. That’s thanks to a micro cloud, with all of their key applications ready to go on a hyper-converged platform.”
STEVE ORRIN | Federal Chief Technology Officer, Intel
Inference and Innovation
Moreover, as this technology advances, new opportunities for innovation arise. Take a security application that records activity via a perpetual video feed. Now, when something changes it’s possible to capture and save only the data that the user needs, which, in turn, reduces the amount of data that has to be saved, processed and stored overall.
“I need the inference, not the data,” Orrin explains. “I need to know that there was a tree there and now there isn’t or that there was a vehicle, but I don’t need to see 12 hours of raw data where nothing’s changed. If I want that sensor platform deployed for longer periods of time, I need to change the requirement for how much data I’m going to save. This is a mental shift to rely more on the intelligence in the platform to do the work rather than having to do it manually.”
By relying on technology to make real-time decisions rather than analysts’ hours, days, or weeks after the fact, agencies can accelerate the speed at which intelligence is recognized and analyzed. That’s where we are today, Orrin says. “It’s a hybrid cloud, an edge cloud, and it’s across multiple clouds,” he notes. “It’s not about having one architecture or the other. It’s about doing the right processing, at the right place, at the right time.”
“It’s not about having one architecture or the other.
It’s about doing the right processing, at the right place, at the right time.”
STEVE ORRIN | Federal Chief Technology Officer, Intel
Adopting a Data-Driven Architecture
To get this right, users need to think hard about the data and consider where the best place to compute, store and process it might be. “You want to be able to do the processing and intelligence gathering as close to the data acquisition as possible to reduce latency, to reduce the cost of transmission, and to reduce the cost of storage,” Orrin says. The right data — but not all the data — needs to return to the backend systems. There, staff can conduct Big Data analytics, leveraging commercial cloud scale, but then extract the intelligence from that data and deliver it to operators, who can then put that information to good use.
“You can call this ‘follow the bit,’ but it’s a data-driven architecture,” Orrin says. “It’s all about the data.
This is what’s driving both the edge cloud and 5G mobile networks: the promise that I can unleash the data-centric applications and devices at the edge to gain a superior picture of the world we live and operate in.”
5G will provide increased bandwidth and speed, enabling more processing on the edge and driving the integration of processed insight on a grander scale.
“With 5G I can increase the number of sensors, because I have the bandwidth within my network, and then I can centralize different mobile edge computing platforms and have them do that edge processing and then deliver the results back to the cloud,” Orrin says. “It’s a more complex, but richer architecture.”
That architecture is powered by Intel technologies. Understanding where Intel sees the future heading is one way federal chief information officers and information architects can ensure they are prepared to enter that future.
“We provide the building blocks for processing, computation, memory, networking and storage for all of these levels of systems from the edge to the cloud, so we bring together unique ecosystems to help drive solutions,” Orrin says. “We can drive integration and interoperability from the get-go. Intel is the common component for the telecom providers, the platform makers, the cloud providers: This is not just providing technology ingredients to each of those steps but helping to drive the ecosystem to collaborate and interoperate. That’s how you get solutions.”
Mission Ready: Powering Innovation from Edge to Cloud
Intel’s new resource center provides resources to government decision-makers and helps the public sector make the most of modernization efforts. The resource center is your one-stop-shop for information on the remote workforce, securing the edge, data modernization and offering insights that help you achieve mission success.
© Intel Corporation. Intel, the Intel logo, and other Intel marks are trademarks of Intel Corporation or its subsidiaries. Other names and brands may be claimed as the property of others. Intel technologies may require enabled hardware, software or service activation. No product or component can be absolutely secure.
