Cloud computing has revolutionised the way we manage and procure compute resources. It’s ubiquity and utility style management have helped make the approach permeate into and transform every industry imaginable.
Combining Artificial Intelligence (AI) with cloud computing, making AI readily accessible and easier to setup and consume has also enabled new approaches and new types of applications.
And when you factor in the Internet of Things (IoT) – connected devices, providing interactivity and vast sources of new data, streamed to and managed from the cloud – enabling insights, features and functions that were once science fiction and are now commonplace, transforming everything from the way airlines buy and manage jet engines, to how consumers secure & heat their homes.
But there are limits.
Centralising your data, even if it’s widely distributed in the cloud, means the devices need to be connected and for some applications – think video monitoring as an example – the bandwidth requirements could be huge. 5G networks might help with that, but even so the cost and capacity limits in a solution where everything needs to be pushed into the ‘middle’ before it can be acted upon are prohibitive.
Add to this the latency issue – even with fast networks, data has to be sent from device to the ‘analyst’ in the cloud, processed, queued and then a response sent back, all introducing lag. In some use cases that might be OK, but if it’s analysis of a video from the camera watching the road in your self-driving car even a fraction of a second’s lag ,or a momentary black spot issue is unacceptable.
That’s where the “Intelligent Edge” comes in.
Microsoft define it (paraphrasing) as a set of connected systems and devices that work together to analyse data close to the users, the data or both. The labels here aren’t as important however as the concept.
This model means pushing some of the intelligence, some of the work, to the ‘edge’ of the system or ‘on/near device’, allowing real-time insights (or action) and providing highly responsive and context-aware apps. Combined with additional services ‘off device’, such as processing, storage, data aggregation, AI and Machine Learning over both the current data and all available historic data in the cloud and you have a paradigm that facilitates the potential for a new class of distributed systems, connecting people, processes and data through new applications that deliver truly breakthrough business outcomes.