Proliferation of the IoT is creating a need for low latency, smart decision-making close to the embedded system
While the current global market for IoT Edge Analytics is relatively small, it is expected to grow at a rapid 5-year CAGR of over 35% through 2022 according to a new report by VDC Research. New business models that encourage cloud consumption over licensing at the edge, in addition to immaturity in this new market have combined to keep current market revenues down. Still, engineers expect to more than double the portion of projects using edge analytics over the next three years.
The sheer volume of raw data generated by IoT systems has made it clear that significant portions of data must be filtered or pre-processed to reduce the volume of data sent to the cloud for cost and latency purposes.
“The current wave of innovation at the edge is due to improvements in IoT and embedded hardware as well as advances in engineers’ understanding of and ability to deploy advanced machine learning (ML) self-training algorithms,” said Roy Murdock, IoT & Embedded Technology Analyst at VDC Research.
“There will always be a need for low latency, smart decision-making as close to the embedded system as possible.”
As the dust settles around hardware at the edge in the gateways market, software vendors are stepping up to add value and differentiation through analytics at the edge.
VDC’s research shows that the energy & utilities market makes up the largest IoT edge analytics vertical, amounting to nearly 40% of global edge analytics revenues in 2017.
“The early and widespread adoption of gateways and other edge hardware in remote environments, such as offshore oil rigs, mines, and power generation plants, has allowed this vertical to maintain its leading spot,” explained Murdock. “In many cases connectivity is expensive and intermittent in these deployments, making them perfect candidates for edge offline syncing and filtering.”
Application container technology is a key component in many IoT edge analytics solutions, according to VDC. The heterogeneous nature of both hardware and software on edge deployments calls for a solution that can scale across these different environments.
Leading vendors are turning to containers to help them solve this issue and create extensible, hardware and OS-agnostic analytics engines and runtimes. VDC recommends that vendors prepare their strategies to bring advanced analytics closer to the edge by developing new micro-container options and building up their custom data science and consulting teams.
