Guest post by Andrew Foster, IOTech Product Director
Twelve months ago, who could have predicted that 2020 would turn out the way it has? As a tech company focused on providing software solutions to the rapidly evolving world of edge computing, we anticipated a bumper year of accelerated adoption and, finally, wide-scale deployment for our edge technologies. Undoubtedly, COVID-19 was a shock to most organizations with many projects delayed or even put on hold while companies scrambled to figure out what the pandemic meant for their business and customers. Our customers’ projects didn’t disappear altogether but were simply delayed. As we approach the end of the year, our users have readjusted to this new world. Edge adoption is moving ahead with some optimism. In fact, out of the COVID-19 crises, new opportunities and use-cases tied to supporting the fight against the pandemic are emerging as an opportunity for the edge computing community.
Despite setbacks at the start of the pandemic, we are still seeing increased interest in edge solutions. A number of trends have emerged during 2020 that point to the future direction of edge adoption should be the focus for vendors in 2021.
From Centralized to Hybrid Architectures
There is currently an acceleration from fully centralized cloud-based systems to distributed architectures driven by edge computing. Cloud computing alone is unable to handle the vast amounts of data that will be created by the billions of connected devices predicted, or the need for local insights from the latency-sensitive applications on which they depend. What is also clear is that most systems will not rely 100 percent on edge computing solutions either. In fact, fully autonomous edge applications are quite rare. Most systems will require a hybrid solution consisting of both edge and cloud components. This is because many users require the ability to utilize cloud resources for heavy-duty applications, while using edge computing for lighter-weight processing and local, real-time insights. In this hybrid model, interoperability between edge architectures and cloud (public and/or private) environments is absolutely key.
Wide Adoption of AI and ML
The use of advanced model-based AI and ML applications at the edge is becoming mainstream. Edge AI supports a new generation of real-time data-driven use cases, including predictive analytics, robotics, autonomous systems and image recognition. AI algorithmic frameworks have matured significantly in recent years. But as we head into 2021, there is still a need for further development in the ability to retrain models, typically in the cloud-based on live operational data, and then to easily automate the re-deployment of new models onto the edge nodes.
Continuing with the theme of AI and ML adoption, video inference has emerged as a “killer” edge application in 2020. There is huge interest from edge computing users for IP camera device integration combined with the ability to analyze video streams in real time for image recognition or object counting. This is driving a broad range of use cases across different vertical markets, ranging from manufacturing, oil and gas to retail. At least in this area of edge computing, COVID-19 has actually created new opportunities for solution providers to create edge solutions. For example, solutions can use the analysis of video streams to help maintain social distancing. They will monitor the distance of individuals from each other in enclosed spaces and automatically detect unsafe behavior. The analysis of thermal camera images by edge systems also supports the automated detection of abnormal body temperature when screening large numbers of people in real time.
Edge Management at Scale
Finally, experience with projects in 2020 has shown the next big challenge to large-scale adoption and rollout of edge computing is determining how to manage edge deployments at scale. At the edge, users face a unique set of management challenges when deploying systems consisting of hundreds or possibly thousands of edge nodes. A cloud native solution, such as Kubernetes, is too heavyweight for many edge environments. What do you do if the edge application doesn’t run in a container, as is the case if you’re deploying an edge application into an environment such as an MCU? In any case, application management is only a part of the problem. The edge nodes themselves also need to be managed and monitored. Connecting and provisioning large numbers of edge devices/sensors that produce the data to the edge compute nodes is also a significant challenge as the scale of a system increases. These challenges are relatively easy to address in small-scale pilots. But as edge computing moves closer to large-scale rollout, the edge needs its own solutions.
As 2020 closes, the business climate is improving, even if the world is not fully out of the pandemic. The world didn’t stop, although for a time, it seemed as if it was stalling. At IOTech, we think 2021 is going to be a great year. We see lots of new opportunities that will emerge, driven by increased edge computing adoption. Big challenges still exist, however. This is what makes the edge computing world so much fun!
Goodbye, 2020. You taught us a lot. Let’s roll on to 2021!
About the Author
Andrew Foster is a Product Director at IOTech and has more than 25 years of experience developing IoT and other Distributed Real-time and Embedded (DRE) software products and services. He has held senior management positions in product delivery, product management and product marketing. Before joining IOTech, he was the Product Marketing Manager at PrismTech and led a team responsible for helping to grow PrismTech’s global market presence. Andrew holds an M.S. in Computer-Based Plant and Process Control and a B.Eng in Digital Systems Engineering.
DISCLAIMER: Guest posts are submitted content. The views expressed in this blog are that of the author, and don’t necessarily reflect the views of Edge Industry Review (EdgeIR.com).
AI/ML | business strategy | COVID-19 | edge application | edge computing | edge device | IOTech Systems