Introduction to AI at the Edge — A Beginner’s Perspective

Before we jump on to Talk About , Edge Computing and AI at the Whole , let’s first talk about the Internet of Things Platform and its Relation with Cloud , so that we can derive differences between the Cloud Computing Platform and Edge Computing Platform .

Let’s first Talk About the Layman Meaning of Internet of Things . If we look up in the Dictionary , IoT is defined as Follows :

Interconnection of Computing Devices embedded in Everyday Things via the Internet

Essentially , it is everyday things connected to the Internet and hence , Internet of Things . Now , in this Path of Interconnection there are 3 interacting Units :

1).There is a Network of Objects that Connected Wired or Wirelessly to the Internet , and these Objects can be anything ranging from Appliances , Cars etc

2).Next , there are Gateways with access bridges between these Devices and the Internet .

3).And , the Final thing is the Cloud where we want to store the Information which we are receiving from the Devices for processing it and then analyzing it to Derive Intelligence .

Now , the Data that is being Generated by these IoT Devices is increasing Exponentially , and the Cloud Platform is not sufficient enough to store this Data . So , we need a substitute for this to combat this problem .

Now , here comes in play the Edge Computing Platform . There is an emergent need for running Local Services close to the Devices itself , to the Edge Devices itself .

Now , lets dwell into What do we mean by this Term “Intelligence at the Edge” . Let’s take some Example Use Cases :

Suppose you have a Camera pointed at your Parking Lot , where you are using the Camera to count in Real Time , the number of Cars that are Parked in the Parking Lot . It isn’t really cost effective and it doesn’t make any sense to send every frame of the Video from the Camera to Cloud to Analyze and count the Number of Cars that are Parked in the Parking Lot .

What if we could do it in the Camera itself ?

If you are using Object Detection using your Camera . For Privacy Reasons , you may not want to send your Data to the Cloud , but you would want to run some of those analyses and intelligence on the Device itself .

So these above two reasons are the major reasons for Using Intelligence at the Edge .