Many companies selling IoT-based edge solutions are now promoting a hybrid approach
Despite the expansion and variety of available edge computing devices, it is unclear whether they will be thin or thick. Edge computing emerged in the thin space. Consider these as low-power, low-memory devices that quickly send data to a centralized system for processing or analysis. The majority of businesses use this method to gather streaming data to provide real-time analytics.
If thin edge devices do act independently, they typically respond to light or motion sensors by performing a straightforward operation. With thin edge devices operating as a “bridge” that securely transports orders and data between local equipment and the public cloud, they have also been widely used in larger businesses as they have attempted to link their legacy or on-premises hardware to their public cloud.
On the other hand, thick edge refers to powerful computing systems that can handle more intensive processing, sophisticated logic, and complicated resolution. Because they can act independently and frequently utilize AI tasks to do so, they could have worse or more erratic connectivity than their thin counterparts. When these devices do transfer data to a central cloud storage location, it’s frequently after the data has been normalized, anonymized, filtered, or otherwise altered to meet a specific requirement.
These gadgets are what you can currently find in plant floor IoT infrastructure that enables on-machine, on-demand sensitivity to challenging issues, like monitoring the health of complex machinery and figuring out the best time to shut down the process, and performing preventative maintenance depending on all the inputs. Or turning off a wind turbine or an oil rig in dangerous weather when communication is at best intermittent and no one is available to make a judgment.
Many companies selling IoT-based edge solutions are now promoting a hybrid approach, in which you utilize a mixture of thin and thick edge devices depending on precisely where you need all the costly computing power and where less is better. This is similar to the middle ground between on-premises and cloud architecture.
Despite the expansion and variety of available edge devices, it is unclear whether they will be thin or thick.
In a study of 2,736 IT leaders and buyers and 5,012 business employees, 35% of participants said they had never heard of edge computing, while 20.6% said they were just “somewhat aware” of its capabilities. The awareness performance of midsize and enterprise firms was somewhat better. However, for those just dipping a toe into the edge computing business, the increased complexity will probably prove more perplexing than immediately beneficial.
Businesses that are considering, or have already implemented, a thin, thick, or hybrid edge infrastructure must also deal with the software developments necessary to fully utilize their new technology.
They won’t ever be able to fully realize the value of all the new real-time data if they substantially invest in thin edge devices that only support historical batch assessment and collect data from onboard sensors and promptly transport it to a centralized data lake in the cloud. On the other hand, they can overspend on bulky edge devices in the hope that they’ll eventually “grow into” the computational power but fail to develop workable machine learning models to fully take advantage of the edge AI advancements they were promised.
It is critical to examine the business value you are seeking to generate before making any decisions when thinking about a new edge/IoT infrastructure, whether that be automation, better customer experiences, the interconnection between people, or something else.
The post Thin vs Thick: Edge Computing Enters the Discussion appeared first on Analytics Insight.