Edge computing pushes applications, data, and computing power from central computing locations like servers to the edge of the Internet—to mobile devices, sensors, and end users that are distributed all over the world.
In traditional computing, processing is performed mainly in centralized locations like local servers and in the cloud. If you need a certain service, like image analysis, a camera might send information about images to the cloud to analyze them. However, with edge computing, processing happens close to the source of the data, like in your cell phone or in a device that’s collecting information. Moving computing to the edge has many benefits: faster response times, data privacy and security, and resiliency if the cloud becomes unavailable.
Enabling operations at the edge is important to a growing number of systems. For example, in environments known as the humanitarian edge—where first responders and other emergency personnel work—edge computing can provide faster updates about weather events, seismic events, or infrastructure failures, where conditions can change at a moment’s notice. Similarly, on the tactical edge—where military personnel execute missions in areas that range from air, to sea, to ground—edge computing can improve intelligence, surveillance, and reconnaissance.
But these environments create new software and engineering challenges, especially as more powerful computing is needed to run artificial intelligence (AI) capabilities for the state-of-the-art analysis that humanitarian and tactical personnel need. Devices used by personnel at the humanitarian and tactical edge are often limited in their computational power, and these edge environments lack connections to networks that could support their functionality. To make matters worse, devices on the tactical edge are often the target of cyber attacks by enemy actors.
To support personnel at the humanitarian and tactical edge, we need to make these devices resilient to attacks, and we need to find a way to support them by improving their connectivity to the cloud in a secure, reliable, and timely way. Figuring out these challenges can transform how we conduct everything from military operations to field research, field medicine, transportation, and environmental analysis.
The deployment of IoT devices to support missions, from the enterprise to the tactical edge, is also pushing the cloud-to-edge continuum. The SEI researches how to automate this process—from build in the cloud to secure deployment at the edge.
Grace Lewis SEI Principal Researcher
Serving the Humanitarian and Tactical Edge
The SEI is conducting applied research to create and transition innovative solutions, principles, and best practices for architecting and developing systems that support teams at the humanitarian and tactical edge. The SEI explores how to engineer systems to support edge computing in places where connectivity is limited and on devices that—due to size, weight, and power (SWAP) limitations—have limited computing resources, such as CPU or GPU power and memory.
To make state-of-the-art analyses available to these resource-limited devices, the SEI explores how to design systems for distributability. Distributed systems split computations or analyses into smaller tasks, which are then completed in different places—or nodes. This solution speeds up processing and reduces the computational load on any single device running a subtask of the larger computation or analysis.
Distributability, however, requires communication between these different nodes, which means they need some way to connect to each other. Most devices only work as intended when the connections between them are uninterrupted, which is often not possible at the humanitarian and tactical edge. To support distributability in these special environments, the SEI investigates approaches for engineering software systems to create nodes that can handle intermittent connectivity. In addition, the nodes must be robust enough to work even when communications are actively denied because of attacks by adversaries or because of natural disasters.
The SEI also studies how to reduce the size of the data that edge devices must process, and we’re creating optimization techniques that bring AI and machine learning (ML) capabilities to the resource-limited edge devices that personnel carry in the field. Often, AI and ML capabilities require sophisticated hardware that is unavailable on edge devices, putting advanced analyses out of reach. The SEI’s solutions, however, will place these state-of-the-art capabilities in the hands of humanitarian and tactical personnel to drive the success of their missions.
What We Offer
Establishing Trust in Disconnected Environments
In this podcast, Grace Lewis presents a solution for establishing trusted identities in disconnected environments based on secure key generation and exchange in the field, as well as an evaluation and implementation of the solution.
In our implementation of cloudlets, applications are statically partitioned into a very thin client that runs on the mobile device and a computation-intensive Server that runs inside a Service VM.
Tactical Cloudlets: Moving Cloud Computing to the Edge
This webinar presents the tactical cloudlet concept and experimentation results for five different cloudlet provisioning mechanisms.
Quality Attribute Concerns for Microservices at the Edge
In this webcast, Marc Novakouski and Grace Lewis reviewed characteristics of edge environments with a focus on architectural qualities.
The Latest from the SEI Blog
Networking at the Tactical and Humanitarian Edge
August 08, 2022 • Blog Post
Marc Novakouski, Jacob Ratzlaff
This blog post details networking challenges in edge environments that stem from uncertainty and solutions to overcome...read
Internet-of-Things (IoT) Security at the Edge
April 04, 2022 • Blog Post
Sebastian Echeverria, Grace Lewis
Assuring the security of any hardware device is a hard problem. In particular, Internet-of-Things (IoT) devices have increasingly been the target of malicious...read
Our Vision for the Future of Edge Computing
The growing deployment of IoT devices to support missions, ranging from enterprise to the tactical edge, is also pushing what is becoming known as the cloud-to-edge continuum. Computing capabilities are pushed from the cloud to edge devices to process data collected at the edge, instead of sending all data back to the cloud for processing, which is inefficient. Automating this process—from build in the cloud to secure deployment at the edge—is the focus of our current and future work in edge computing.
The SEI is also investigating how to make AI-enabled data pipelines more resilient. Our solution will monitor components within processing pipelines to detect faults. When faults are detected, the system is reconstituted using substitute components, deploying components on substitute hardware, or adjusting the types of processing within a pipeline. These adaptations ensure continued mission support. Other work to ensure mission support will include designing, developing, and deploying edge-enabled microservice architectures that promote quality attributes such as network resilience, performance, survivability, adaptability, openness, and scalability to adapt to mission needs.
To stay up to date on the future of edge computing, subscribe to our blog or contact us.