Close
Resources

What is edge computing? Why does it matter?

At Teknoir, we often discuss things like “Edge Computing,” and “Artificial Intelligence at the Edge.” This article provides the background information you need to determine how the “Edge” can benefit your organization.

Important Takeaways

  • Edge Computing significantly reduces monthly spending on the Cloud and data plans.
  • Edge Computing removes the need to transmit and receive sensitive data to/from the Cloud, enhancing security and privacy while expediting data processing.
  • Edge Computing runs in places with limited internet connectivity, allowing you to tap into valuable new data streams.

What is Edge Computing?


To understand Edge Computing, it is helpful to compare it with Cloud Computing, which requires knowledge about the “Cloud” and its role in technology and business.

What is the “Cloud?”

Tech and non-tech companies rely on the Cloud to store and analyze data, host websites and software platforms, and create innovative technologies. If you are storing data in Google Drive, reading content on social media platforms, or using streaming services like Netflix, you are sending and receiving data from the Cloud. When data is received by the Cloud, it is either processed (a.k.a., Cloud Computing), stored (e.g., in Cloud Storage), or both.

Behind the scenes, “Cloud” refers to physical computing and storage hardware and the software required to use it. Some examples of Cloud computing and storage providers are Amazon Web Services, Microsoft Azure, and Google Cloud Platform. Each offers pay-as-you-go computers, databases, and managed services that underpin technologies that drive many of today’s businesses forward.

What is the “Edge?”


While many companies have transitioned to the Cloud, this is not always practical.

Certain applications cannot access the Cloud because they operate at the outer boundary of a Wi-Fi, cellular, or satellite network (a.k.a. the “Far Edge”). Deploying software in these settings requires shifting data processing from the Cloud to connected devices that can be deployed in the field (a.k.a. “Edge Computing”).

Shifting processing from the Cloud to deployed devices also enables valuable applications, even in less remote settings with better internet connectivity (a.k.a. the “Edge”). A key benefit is not having to send data from sensors and cameras to the Cloud and wait for a response containing results. In other words, edge computing enables real-time data processing and analysis.

Compared to transmitting everything to the Cloud, collecting and processing data on devices at the Edge enables you to optimize your monthly spending. Just imagine if you owned thousands of smartphones and streamed Netflix on each of them all day, every day. Your wireless bill would be untenable if you pay per gigabyte of data used by each phone, which is typical of most wireless plans for connected edge computers (a.k.a. “Internet of Things” (IoT) devices).

In both Edge and Far Edge scenarios, connected devices are typically owned by the company using them or leased by a third-party vendor. This gives buyers the flexibility to either capitalize their costs by purchasing the hardware upfront or expensing it on a recurring basis.

When should you use Edge Computing?


Edge Computing is important because it decentralizes data processing. When combined with Artificial Intelligence (AI, a.k.a. “Edge AI”) it can automate mundane, repetitive, time-consuming, expensive, and dangerous tasks without human intervention and provide decision support for processes requiring human intervention. Here are three example applications:

1. Real-time Video Analytics to Enhance Workplace Safety

 

Use Case:
Imagine that you are the owner of a construction company. You and your employees want to ensure everyone returns home safely each day. You have implemented administrative safety policies and incorporated physical prevention measures, and you also require your employees to wear personal protective equipment (PPE). Now you want to (a) ensure your employees adhere to the rules for their safety and (b) send supervisors alerts about potential and active incidents.

Recommendation:
Deploy video cameras with real-time Edge AI to monitor PPE usage, identify job site hazards, detect improper equipment usage, and alert when policy violations occur. Teknoir offers customizable and easy-to-deploy Edge AI applications, dashboards, and notifications to enhance workplace safety.

Key Benefit:
Decentralizing AI by moving it to the Edge means we can generate real-time predictive and prescriptive feedback.

2. PPE Monitoring with Employee Privacy Preservation

 

Use Case:
Imagine you are the owner of a manufacturing facility. You want to protect your employees by ensuring they wear PPE while operating dangerous equipment. However, in addition to safety, you and your employees are also concerned about privacy.

Recommendation:
Deploy cameras with an Edge AI solution that offers built-in privacy protection measures. With this approach, you can detect PPE usage around your equipment, automatically alert employees about policy violations, and send images of detected incidents to the Cloud with faces blurred to avoid violating employee privacy. These events can be counted and reviewed from time to time to track and continually improve workplace safety.

Key Benefit:
With Edge AI, it is possible to avoid transmitting sensitive data over the network, improving privacy and security.

3. Monitoring Remote Equipment

 

Use Case:
Suppose you are operating a remote oil rig and need to monitor whether a set of motors is running properly. Production of this rig is relatively low, so you want to minimize the time you spend traveling to and from this site. You also want to avoid paying for an expensive cellular data plan. You only to be alerted when equipment goes offline or needs maintenance.

Recommendation:
Consider deploying vibration and temperature monitoring devices onto each motor with an edge computer that monitors the vibration patterns of each motor. Incorporate Edge AI to detect when sensor readouts deviate from their usual patterns, which can indicate wear and tear and signal maintenance needs. Rather than sending all sensor data over the cellular network, keep cellular data usage low by sending alerts only when motors behave abnormally.

Key Benefit:
Because Edge AI does not rely on sending data to the Cloud, it can be used in places with slow, spotty, or no internet connection.
How to get started

At Teknoir, our mission is to make Edge AI as simple and easy as possible to develop, deploy, and manage. We are excited about this game-changing technology and see new applications appearing daily across every industry.

Contact us to learn more about how we can help you leverage Edge AI to generate value for your business!

Author: Michael Bell, PhD