Edge Computing

Edge Computing

Established: October 29, 2008

Publications

News & features

News & features

Overview

Edge computing is where compute resources, ranging from credit-card-size computers to micro data centers, are placed closer to information-generation sources, to reduce network latency and bandwidth usage generally associated with cloud computing. Edge computing ensures continuation of service and operation despite intermittent cloud connections. Industries ranging from manufacturing to healthcare are eager to develop real-time control systems that use machine learning and artificial intelligence to improve efficiencies and reduce cost. We are exploring this new computing paradigm by identifying and addressing emerging technology and business model challenges.

A Brief History of Edge Computing

On October 29, 2008, we invited colleagues from academia and industry for a day-long brainstorming session about the future of cloud computing. Edge computing was conceived during those discussions. Attendees included Victor Bahl (organizer, Microsoft Research), Ramón Cáceres (AT&T Labs), Nigel Davies (Lancaster University, U.K.), Mahadev Satyanarayanan (Carnegie Mellon University), and Roy Want (Intel Research). Following this meeting, we published the first paper on this topic, in IEEE Pervasive Computing (November 1, 2009) titled: The Case for VM-based Cloudlets in Mobile Computing.

The blog Why a Cloudlet Beats the Cloud for Mobile Apps (December 13, 2009) was the first article to cover our ideas. In it are described two projects, Cloudlets, a joint project of Microsoft and Carnegie Mellon University; and MAUI (Mobile Assistance Using Infrastructure), a Microsoft Research project. In Cloudlets, we investigated fast virtual machine (VM) synthesis on the edge; in MAUI, we explored a .NET programming model for computational offloads to the edge. Many of the ideas we explored have withstood the test of time. For example, guarding against network disconnections, incorporating computing versus communications tradeoff, deciding which methods to offload and which to process locally. The papers describing Cloudlet and MAUI have been cited over 5,700 times. Click here for a colorful description of a ten year look-back

Since then, having made the case for edge computing in the research community (see Faculty Summit keynote), industry (see: Network World interview) and internally in Microsoft (see Intelligent Edge), we have been focusing on live-video analytics as the “killer” app for edge computing. You can read all about it in a separate project page.


Listen to Victor Bahl’s Podcast,  A brief history of networking (and a bit about the future too),
where he shares some fascinating stories and gives an inside look at Edge Computing.


The Intelligent Edge

Microsoft product groups coined the term The Intelligent Edge. The Intelligent Edge is a capability that enables Microsoft customers to enjoy a seamless experience and compute capabilities wherever their data exists—in the cloud or offline. Microsoft is making it easier for developers to build apps that use edge technology, by open sourcing the Azure IoT Edge Runtime, which allows customers to modify the runtime and customize applications.

Types of Edges

Learn more about the Intelligent Edge.

Recent Activity

People

Researchers & Engineers

Interns

  • Portrait of Zephyr Yao

    Zephyr Yao

    UC Irvine

    Summer 2018 | Azure IoT Edge availability using Docker Swarms

  • Portrait of Enrique  Saurez Apuy

    Enrique Saurez Apuy

    Intern - 2017

    Georgia Tech

  • Portrait of Jack Kolb

    Jack Kolb

    UC Berkeley

    Summer 2017 | Declarative specifications for distributed IoT applications

  • Portrait of Shadi Noghabi

    Shadi Noghabi

    UIUC

    Summer 2016, Summer 2017 | Monitoring Azure IoT Edge & software offloading to the Azure cloud

  • Portrait of Chien-Chun Hung

    Chien-Chun Hung

    USC

    2016-2017 | Query optimizing for edge based streaming video analytics system

  • Portrait of Giulio Grassi

    Giulio Grassi

    Intern - 2015, 2016

    Sorbonne Université / LIP6

  • Portrait of Kiryong Ha

    Kiryong Ha

    CMU

    Summer 2014 | GPU state migration between edges and data centers by reproducing OpenGL states

  • Portrait of Aakanksha Chowdhery

    Aakanksha Chowdhery

    Stanford

    Summer 2014 | Edge-based wireless video surveillance

  • Portrait of Tiffany Chen

    Tiffany Chen

    MIT

    Summer 2013 | Vision analytics in real-time with cloud offloads

  • Portrait of Eduardo Cuero

    Eduardo Cuero

    Duke

    Summer 2009, Summer 2011 | Automatic cloud offloading & cloud augmented high-quality gaming on SmartPhones

Collaborators

Research

Applications

From the very beginning, we have maintained that the most compelling applications for edge computing are ones that require low latency responses or ones where the network to the cloud is expensive or inadequate. In this context, we asserted that the “killer app” for edge computing is live video analytics. Along the way, other Microsoft researchers discovered precision agriculture to be a beautiful edge computing application as well. We are exploring both:

Live Video Analytics

Live Video Analytics Scenrios

Large-scale video processing is a grand challenge representing an important frontier for analytics, what with videos from factory floors, traffic intersections, police vehicles, and retail shops. Read more.

Data Driven Agriculture

Photography depicts Microsoft's FarmBeats technology uses AI and IoT to help increase farm productivity.

We believe that data, coupled with the farmer’s knowledge and intuition, can help increase farm productivity and help reduce costs. However, getting data from the farm is difficult since there is often no power in the field… Read more.

Highly Adaptive and Resilient Edges

Continuity of service is a must-have attribute in mission- and safety- critical edge computing applications. For example, in telecommunication networks interruption of communication services is unexceptionable; in oil rigs (link) where continuous monitoring of the safety of on-site workers and the health of multi-million-dollar equipment is a must have; in manufacturing, constantly looking-out for production errors that may lead to defective items is also a must-have. Downtime of edge computing servers, beyond an acceptable threshold, can result in accidents and significant financial loss. The challenge then is to keep everything operational, even when local technicians are absent. This requires edge servers to be up and running (24x7x365), which is particularly hard when edge-cloud connectivity is unreliable and expensive. We are developing adaptive and resilient software solutions that enable edge clusters to run continuously.

Paya

to describe Project Paya

Paya is a state migration edge-tailored solution that ensures that high-availability is met for all edge-cloud applications based on the application’s specific need. Read more.

LOKI

to describe Project LOKI

LOKI is a suite of services and programming abstractions that simplify the development of adaptive edge-cloud & multi-cloud applications, Read more.

Our goal is to develop a software-based system and architecture that can help keep operations alive by healing the system automatically in the face of machine failures, network disconnections, dynamic application loads, or changes in capacity. Towards this north star, we have several ongoing projects that take on these problems.

HybridKube

to describe HybridKube

HybridKube is a Kubernetes extension for optimal placement of applications across a edge-cloud environment. Read more.

Offloading Computations

We have been exploring the fundamental trade-off between computation and communications to enable a new class of cpu-, gpu- and data-intensive applications that seamlessly augment the cognitive abilities of users by exploiting speech recognition, NLP, vision, machine learning, and augmented reality (Project Maui, Mobisys 2010). We have made significant progress in overcoming the energy and computation limitations of sensors, handhelds, and wearables. In subsequent research we demonstrated how important special-purpose workloads can also leverage cloud offload: for GPU-intensive rendering applications (Project Kahawai, MobiSys 2015) and deep neural network video stream processing (MCDNN, MobiSys 2016).

Project MAUI

Image of chess being played on phone to describe Project Maui

Mobile Assistance Using Infrastructure (MAUI) was the first system to demonstrate fine-grained code offload to nearby edge server(s) with minimal programmer effort. Watch the video.

Project Kahawai

Image of person on computer

Kahawai enables high-quality gaming on mobile devices, such as tablets and smartphones, by offloading a portion of the GPU computation to server-side infrastructure. Watch the video.

Geo-distributed Edge Analytics

Edge servers located in thousands of locations and managed by the same administrative entity offer powerful computing resources for cloud providers. Our research on low-latency edge analytics explores how best to use these resources. For example, the old approach of aggregating all the data from sensors to a single data center negatively impacts the timeliness of the analytics. But, running queries over geo-distributed inputs using the current intra-DC analytics frameworks also results in high query response times because these frameworks cannot cope with the relatively low and variable capacity of the WAN links. Our Iridium system (SIGCOMM 2015) provides low latency geo-distributed analytics by optimizing placement of both data and tasks of the queries. Follow-on work (CLARINET, OSDI 2016) considers WAN links with heterogeneous and modest bandwidths, unlike intra-datacenter networks, when deriving query execution plans across the cloud and edge servers.

ML for Edge

Our colleagues in Microsoft Research India are developing a library of efficient machine learning (ML) algorithms that can run on resource-constrained edge and IoT devices ranging from the Arduino to the Raspberry Pi. The thesis is that IoT devices and sensors don’t have to be “dumb” i.e. they can do more than just sense their environment and transmit their readings to the cloud, which is where the traditional decision making intelligence resides. Instead, an alternative paradigm is where even tiny, resource-constrained IoT devices run ML algorithms locally. This enables important scenarios overcoming concerns around connectivity to the cloud, latency, energy, privacy, and security. Read more

Networking

Cloud providers, such as Microsoft, have two types of edges: On-net edges or Off-net edges. On-net edges are generally easier to operate, manage and maintain as they are on the cloud provider’s network. In contrast, Off-net edges are connected to the cloud via the Internet, which may include several ISPs. Managing and operating such edges can be challenging due to the vagaries of the Internet. We are investigating problems to improve the network connectivity to our Off-net edges and the networking between the edges and sensors. Furthermore, edges provide us an opportunity to (re) investigate old ideas around low-latency, secure, overlay networking.

Security

Cloud companies spend large amounts of money to physically secure their millions of servers located in their many data centers. In contrast, edge computing servers may or may not be physically secured. This opens the possibility of malicious attacks on the edge and cloud infrastructure. While a lot has been done to physically secure assets in the cloud, we are investigating techniques to do the same for our edge assets. Security and trust require authenticity and integrity, so we are investigating the use of sensors and specialized hardware in combination with new programming abstractions and system support for building secure and trusted edges. This research builds on our prior work on trusted sensors (MobiSys 2012, ASPLOS ’14) and recent product offering (Azure Sphere).

Cloud Services

Before the dawn of edge computing, which has brought about a major cloud computing paradigm shift in the industry, we developed, deployed and operated a cloud service-store under the banner of Project Hawaii. With it we empowered developers to build sophisticated, cloud-enhanced applications for their resource constraint devices. Our cloud service store included a variety of services including: optical character recognition, speech-to-text, path prediction, social computing, language translation, relay, rendezvous, etc. for Windows, Android, & IOS devices. Over 60 universities used our services as a teaching aid for senior and graduate-level mobile + cloud computing courses. 2015 onward similar cloud services were commercialized by all major cloud providers under the banner of cognitive services. Check out Azure cognitive services. Historically speaking, Project Hawaii was the first to show how cloud/edge can be used in conjunction with a resource-constraint mobile device to augment human abilities.

Project Hawaii

Project Hawaii group photo

The Project Hawaii team – BACK ROW (left to right): Gleb Krivosheev, Philip Fawcett, Ronnie Chaiken; FRONT ROW (left to right): Arjmand Samuel, Jitu Padhye, Alec Wolman, Victor Bahl. Read more.

Gallery

Gallery image of Project Hawaii

A utility tool developed by a student for on-the-go translations. Project Hawaii’s OCR & S2T services, and Bing Translator were used. Check out our Gallery for dozens of student created featured projects. Read more.

Keynotes

Mar. 5, 2020 | Future of Information and Communication Conference (FICC) 2020 – Edge Computing for the (Telecom) Infrastructure | Victor Bahl | Download Presentation

Nov. 11, 2019 | IEEE International Conference on Industrial Internet 2019 – Edge Computing: Where are we today and what’s next? | Victor Bahl

Oct. 3, 2019 | CRA/CCC Visioning Activity: Wide Area Data Analytics – Live Video Analytics (extracting actionable insights from cameras) | Victor Bahl | Download Presentation

Aug. 5, 2019 | The 1st International Workshop on Artificial Intelligence of Things – Fueling Industry 4.0 with the Intelligent Edge | Victor Bahl

Jun. 12, 2019 | 20th IEEE Symposium on a World of Wireless, Mobile and Multimedia Networks 2019 – Edge + Wireless: Technologies fueling Industry 4.0 | Victor Bahl

April 16, 2019 | Cyber-Physical Systems and Internet-of-Things Week 2019 – Edge computing and the fourth industrial revolution | Victor Bahl

February 6, 2019 | Fourteenth Annual University of Illinois Urbana Champaign CSL Student Conference – Better together: the intelligent edge + the intelligent cloud | Victor Bahl

August 20, 2018 | SIGCOMM Workshop on Big Data Analytics and Machine Learning – Democratizing Video Analytics – The quest for the holy trinity of low latency, low cost, and high accuracy | Ganesh Ananthanarayanan

August 20, 2018 | SIGCOMM Workshop on Mobile Edge CommunicationsEdge computing: a historical perspective & direction (slides) | Victor Bahl

July 27, 2018 | Bleeding Edge of Intelligent Edge – Edge computing: 10 years and counting (video) | Victor Bahl

October 23, 2017 | IEEE Fourteenth International Conference on Mobile Ad Hoc and Sensor SystemsLive Video Analytics (video) | Victor Bahl

October 15, 2017 | Third IEEE International Conference on Collaboration and Internet Computing – Democratizing Video Analytics | Victor Bahl

September 29, 2017 | Emerging Topics in Computing Symposium, University at Buffalo Computer Systems Engineering Dept. 50th Anniversary – Live Video Analytics the Perfect Edge Computing Application | Victor Bahl

December 10, 2016 | IEEE International Performance Computing and Communications ConferenceDemocratization of Streaming Video Analytics & the Emergence of Edge Computing (video) | Victor Bahl

May 13, 2015 | Devices and Networking Summit 2015Cloud 2020: Emergence of Micro Data Centers for Latency Sensitive Computing | Victory Bahl

March 10, 2015 | IEEE Wireless Communications and Networking Conference (WCNC) 2015Cloud 2020: Emergence of Micro Data Centers (Cloudlets/Edges) for Latency Sensitive Computing (slides) | Victor Bahl

February 19, 2015 | IEEE International Conference on Computing, Networking and Communications (ICNC) 2015Cloud 2020: Emergence of Micro Data Centers (Cloudlets/Edges) for Latency Sensitive Computing (slides) | Victor Bahl

June 27, 2014 | MSR Summer School on Advances in Wireless NetworkingCloudlets for mobile computing (slides) | Victor Bahl

November 22, 2013 | 2nd IEEE International Conference on Cloud Networking (Cloudnet) 2013 – Cloud 2020: Emergence of Micro Data Centers for Latency Sensitive Computing | Victor Bahl

Additional news

How we created edge computing

Edge computing processes data on infrastructure that is located close to the point of data creation. Mahadev Satyanarayanan recounts how recognition of the potential limitations of centralized, cloud-based processing led to this new approach to computing.

Nature Electronics | January 16, 2019


Microsoft Azure enables a new wave of edge computing. Here’s how.

We are going through a technology transformation that is unlocking new scenarios that were simply not possible before. Smart sensors and connected devices are breathing new life into industrial equipment from factories to farms, smart cities to homes, while new devices are increasingly cloud connected by default – whether it’s a car or a refrigerator.

Microsoft Azure Blog | September 24, 2018

Microsoft will invest $5 billion in IoT. Here’s why.

Today, we are announcing that we will invest $5 billion in the Internet of Things over the next four years. The reason we are doing this is simple: Our goal is to give every customer the ability to transform their businesses, and the world at large, with connected solutions.

Microsoft IoT Blog | April 4, 2018

Could AI end car accidents?

Imagine for a moment that, every week, four to five commercial airplanes crashed in America. In reality, a similar number of people die per week in traffic accidents, but, for the most part, those deaths don’t resonate with us in the same way.

Technical.ly | September 6, 2017

Microsoft Build 2017 buzzword bingo: On the edge

What Microsoft is doing in the edge computing/distributed computing space may be a big topic at this week’s Build 2017 developer conference. Here are a few clues as to why.

ZDNet | September 6, 2017

Why a Cloudlet Beats the Cloud for Mobile Apps

Sure, you know cloud computing. You also know a bit about so-called “private clouds,” which enterprises and government agencies are exploring as an option to combine the power and scale of virtualized cloud architectures with security and control over data. But what do you know of Cloudlets? They may just be a key to the future of mobile computing.

Shepherd’s Pi | December 13, 2009


The Case for VM-based Cloudlets in Mobile Computing

Resource poverty is a fundamental constraint that severely limits the class of applications that can be run on mobile devices. This constraint is not just a temporary limitation of current technology, but is intrinsic to mobility. In this paper, we put forth a vision of mobile computing that breaks free of this fundamental constraint.

IEEE Pervasive Computing | November 1, 2009

Engagements

Events (we helped organize)

Research Community Service

Special Issue

2020 | Guest Editor (Victor Bahl): Call for Papers: Pervasive Computing at the Edge, IEEE Pervasive Magazine

Journals

2017 – Present | Associate Editor (Victor Bahl): ACM Transactions on Internet of Things
2017 – Present | Associate Editor (Victor Bahl): IEEE Transactions on Service Computing
2013 – Present | Advisory Board Member (Victor Bahl): IEEE Internet of Things Journal
2007-2018 | Editorial Board Member (
Victor Bahl): Foundations and Trends® in Networking

Conferences & Workshops

2019 | Program Committee Co-Chair (Ganesh Ananthanarayanan): The Fourth ACM/IEEE Symposium on Edge Computing
2018 | Program Committee Co-Chair (Victor Bahl): The Third ACM/IEEE Symposium on Edge Computing
2018 | Invited Speaker (Ganesh Ananthanarayanan): IEEE Sarnoff Symposium
2018 | Program Committee Co-Chair (Ganesh Ananthanarayanan<): 10th USENIX Workshop on Hot Topics in Cloud Computing
2016 | Advisor & Steering Committee Member (Victor Bahl): NSF Workshop on Grand Challenges in Edge Computing
2014 – Present | (Founding) Steering Committee Member (Victor Bahl): ACM/IEEE Symposium on Edge Computing
2010-2015 | (Founding) Steering Committee Member (Victor Bahl): ACM workshop on Mobile Cloud Computing and Services (MCS)

Distinguished Seminars (on Edge Computing)

September 2018 | University of Southern California, Los Angles, CA | Ganesh Ananthanarayanan
September 11, 2017 | Rice University, Houston, Texas | Victor Bahl
April 28, 2017 | Washington University St. Louis, St. Louis, Missouri | Victor Bahl
December 17, 2014 | Sorbonne Université, Paris, France | Victor Bahl
November 20, 2014 | University College of London, London, U.K.| Victor Bahl
October 3, 2014 | Yale University, New Haven, Connecticut | Victor Bahl

Panels (on Edge Computing)

November 7, 2019 | Edge Computing: Where are we today and what’s next? | Victor Bahl
February 19, 2019 | AI/ML for Communication Networks | IEEE Intl. Conf. on Computing, Networking & Communication | Honolulu, Hawaii, USA | Panelist: Victor Bahl
October 12, 2017 | Enabling Technologies for Edge Computing | Second ACM/IEEE Symposium on Edge Computing | San Jose, California, USA | Panelist: Victor Bahl

Outreach

The Intelligent Edge is a continually expanding set of connected systems and devices that gather and analyze data—close to your users, the data, or both.

Microsoft product groups coined the term The Intelligent Edge. The Intelligent Edge is a capability that enables Microsoft customers to enjoy a seamless experience and compute capabilities wherever their data exists—in the cloud or offline. Microsoft is making it easier for developers to build apps that use edge technology, by open sourcing the Azure IoT Edge Runtime, which allows customers to modify the runtime and customize applications.

Collaborations

Video Analytics over 5G Project

Princeton University

Internet-enabled cameras pervade daily life, generating a huge amount of data, but most of the video they generate is transmitted over wires and analyzed offline with a human in the loop.  As a result, the amount of coverage and level of application accuracy that today’s surveillance camera systems can provide is necessarily limited.  Work has commenced both on scaling the stream processing behind video analytic systems and leveraging certain aspects of 5G technologies such as small cells.  These ideas have in turn enabled exciting new applications for computing on the edge.  The maturation of deep learning techniques and the complete 5G portfolio of technologies create exciting new opportunities to tackle even more challenging problems in video analytics.

This research is centered around Live Video Analytics  occurring over a 5G network with multiple cameras and research program that leverages (a) The full suite of 5G technologies, including rapid mobility handover, small cells, and millimeter-wave (24 and 60 GHz) radio frequencies, and (b) Massive arrays of video cameras, backed by deep learning algorithms to process video jointly across the entire array of cameras.    The use of 5G wireless links to each camera enables an unprecedented amount of wireless capacity to the edge devices, enabling buffering to be relegated to the edge device rather than situated onboard the camera

Living Edge Lab

Carnegie Mellon

Carnegie Mellon University and Microsoft are collaborating on a joint effort to innovate in edge computing, an exciting field of research for intensive computing applications that require rapid response times in remote and low-connectivity environments.

By bringing artificial intelligence to the “edge,” devices such as connected vehicles, drones or factory equipment are able to quickly learn and respond to their environments, which is critical in scenarios like search and rescue, disaster recovery and safety.

To enable discovery in these areas and more, Microsoft will contribute edge computing products to Carnegie Mellon for use in its Living Edge Laboratory, a testbed for exploring applications that generate large data volumes and require intense processing with near-instantaneous response times.