With the explosion of the data which are being generated and processed in an Internet of Things (IoT) system, Cloud Computing approach would be very costly and inefficient due to the cloud processing latency. Such situation has motivated the growth of the Edge Computing approach in IoT ecosystem resulting in more and more IoT developer start to explore and evolve the centralised processing to distributed architecture.
For some time-critical application, the edge processing approach helps a lot in terms of latency, some minor signal processing could be done in the IoT devices instead of processing most of them in the cloud and then only perform respective actions to the devices. By utilising the excessive processing capacity in the microprocessor, some of the minor or simple decisions can actually be made at the device level, which directly reduced the processing load in the cloud and make the device to be more responsive.
Storage is another costly resource for IoT and Big Data application. Since not all the collected raw data are needed for analytics, some data can actually be filtered and processed at the device level. This will allow creation of meaningful data before sending to the cloud computing unit to analysed. Hence, the database hosting cost can be reduced with lower size of the database, and reduce the cloud computing load.
For instance, there is the recently announced Nest Cam IQ indoor security camera, which uses on-device vision processing to watch for motion, distinguish family members, and send alerts only if someone is not recognized or doesn’t fit pre-defined parameters. By performing computer vision tasks within the camera, Nest reduces the amount of bandwidth, cloud processing, and cloud storage used versus the alternative of sending raw streams of video over the network. In addition, on-device processing improves the speed of alerts while reducing chances of annoying, recurrent false alarms. (extract from source)
Offline response might be crucial for some of the IoT applications, on-device processing can improve the user experience when it still can functionable without internet access, since those data are locally saved and locally processed.
Security and privacy can also be improved with edge computing by keeping sensitive data within the device. Edge computing can reduce the amount of devices connected to the internet, it reduces the potential cyber attack to the system. Local data processing at the edge devices can also reduce the amount of sensitive and private information to be sent over the internet.
