Primary goal of any IoT system is to get the useful insights from the data it generate. Edge computing is getting popular due to its low latency and other benefits over cloud computing. Cloud and analytics are inseparable part of an IoT system.
Definition:
According to Sisence “Edge analytics is a model of data analysis where incoming data streams are analyzed at a non-central point in the system such as a switch, a peripheral node, or a connected device or sensor “
Traditional way of analytics:
- capturing from generating sources
- storing in database
- operating stored data
- displaying the result
The problem with this way of building and doing analytics is the data travels a lot. These different steps occur in different places. For example, data is captured from a mobile phone in your hand, is streamed thousand of kilometers to a server to be written to a database, and the result is streamed back.
How edge analytics solves above problem?
The concept of edge analytics solves this problem, by doing as much of the analytics process as possible in devices (aka ‘near the edge’ of the process) to reduce the amount of data travelling.
Rather than streaming a few terabytes of data from a traffic camera to a remote server and database, a machine learning system can be deployed in the camera itself to intelligently spot (most) anomalies and only capture those. This vastly reduces the amount of data travelling, and promotes capturing what is interesting rather than blindly capturing everything.
Advantages:
- Lowest possible latency till today
- Decreased storage and operational cost
- Linear scalability of IoT devices
How critical is Edge analytics?
When we consider sensitive systems like Medical systems , Smart homes , Electric vehicles where latency matter a lot than anything else . In those systems we cannot wait for small amount of delay. That’s why 5G technology is much essential booster for an IoT system to work smoothly.With 4G networks, you’re looking at an average latency of around 50ms. That could drop to 1ms with 5G technology !!!.
So can edge analytics replace cloud analytics completely?
Honestly, edge analytics is not here to replace cloud analytics completely, but it is here to complement cloud analytics by driving near real-time analytics as it is close to the data source.
Market of edge analytics:
- It is gaining increasing demand owing to the emergence and the booming expansion of the Internet of Things (IoT) and the fast-paced growth in the availability of data through connected devices and via real-time intelligence.
- The manufacturing industry can make extensive use of it , for instance, in a smart production line. Pointing out manufacturing defects or anomalies, badly printed stickers, packaging, etc., in real-time can be achieved .
- Healthcare is another domain where a massive surge in the number of connected devices can be witnessed. A large hospital can have as much as 85,000 connected medical and IoT devices, putting a significant strain on the cloud network.
Major companies in this field:
Cons :
Edge analytics is not a silver bullet – by capturing less (Edge computing not all data is used only important data is used )we might miss some things. But having played around with sensors and IoT data, the sheer volume means simply storing everything is pretty unattractive as well. So I think the concept of edge analytics will be a good option to cope with the deluge of data.
Conclusion:
The broader principle still remains – different analytics architecture for different use cases depending on the problem you are trying to solve.