In the early days of modern Internet of Things systems, it was often possible to push all data to a central cloud for processing. This is still true today for many IoT pilots and POCs. But what many organizations realize, sooner or later, is that what works at lower volumes of data is no longer workable or cost-effective at higher volumes. Google Trends gives us some data to back this up:
Beginning in 2014, searches for IoT and machine learning began to surge. This was followed approximately 3 years later by a surge in edge computing searches (which is still continuing today). Clearly, those early implementors of IoT and AI systems began to hit some kind of wall for how to process high volumes of data, and are searching for solutions.
AI and machine learning is an area where the ability to process data close to where it is generated is becoming very critical. Central clouds are well-suited for model training but to push gigabytes of video, audio, and raw machine telemetry data to the cloud to run models against, frankly makes no sense.
Then there are issues such as:
- Data privacy: healthcare clients are not able to push hospital sensor data outside of the hospital
- Data security: electrical grid clients are prohibited by law from opening their systems to the public internet
- Connectivity: energy clients operating in difficult environments need to keep their applications and systems running whether or not they have a connection to central systems
For all these reasons, and more, effective IoT (and AI) simply requires the edge. Happily, Pratexo makes orchestrating and automating the design, deployment, and management of multi-tier, distributed computing infrastructures all the way to the far edge as easy as cloud computing is today.
