In this blog, I’ll define edge computing, why it’s important, and share some examples of where the edge plays a key role in automation and autonomy.  

What is Edge Computing?

Edge computing brings the processing of data close to the things and people that are producing and consuming the data.

The term “edge” is almost synonymous with “local”, referring to the data at the network’s edge – actually on, near or around the physical thing producing the data, such as a machine, networking tower, autonomous vehicle or robot. Edge computing uses embedded computers, modules, and integrated hardware/software systems that allow the data to be extracted, computed and shared at the source.

Why Does Edge Computing Matter?

The edge is where business value is created. Edge computing is more efficient and cost-effective at turning raw data into valuable control data that is actionable, which is critical for a future where machines are becoming more automated, and more autonomous.

Gartner predicts that by 2025, 75 percent of enterprises’ data will be created and processed outside the data center or the cloud. Additionally, ABI Research anticipates that 43 percent of artificial intelligence tasks will occur on edge devices by 2023.

Bringing intelligence closer to machines, where data is created, is a key factor in the efficiency and effectiveness of a high-demanding AI and IoT systems, or AIoT. It removes points of failure from critical decision-making and operations, as well as improving latency and costs.

How Will AI be Used with Edge Computing?

Edge computing is the catalyst for AI, hence ADLINK’s vision statement. Bringing AI to the edge expands the number of ways businesses can gain value from AI because it allows data to be acted upon when it’s still valuable, as it’s being produced, in real-time. For instance, there is little value of detecting a fault on a production line an hour after its occurred.

We’ve noticed a large percentage of industrial data and business data typically is not processed because of the time value of data – by the time data would normally be processed to drive an outcome the data would no longer be relevant. This is where edge AI is a game-changer:

• Edge AI and networking – there is a reason Multi-access Edge Computing (MEC) and the advancement of 5G go hand-in-hand. Bringing AI to the networks edge allows for faster AI computing, reduction in network latency and associated costs. Train AI models in the cloud, run them at the edge is a motto we like.
• Edge AI and fault tolerance – paraphrasing Murphy’s Law, “anything that can go wrong will go wrong”- here at ADLINK, our edge AI solutions are built modularly. In the unforeseen event that a component fails, the rest of the solution still works, operations still continue. With Edge AI there is also the ability to run AI inference offline if needed.
• Edge AI and privacy – often times, for legal reasons, data cannot leave the premises. Edge computing allows data to remain local.

What Edge AI Applications are we Seeing?

Applications already exist for edge AI and continue to expand. Any business with mission-critical operations can benefit from edge computing.

• Manufacturing & distribution: it’s critical that production lines operate with the highest quality, and fulfillment accuracy is correct when it comes to operational costs, customer satisfaction and employee ergonomics. Edge AI and machine vision is ensuring packing and pallet accuracy within warehouse fulfillment and automating arc welding quality at John Deere.

• Robotics and autonomous vehicles: safety, reliability and interoperability is critical in robotics and autonomous vehicle development and deployment. Edge AI is currently powering autonomous vehicles in the world’s first head-to-head high-speed autonomous land race in the Indy Autonomous Challenge.

• Energy and utilities: worker safety is critical in the energy and utilities industry, sometimes all it takes is one wrong move without proper protective equipment for an employee to require medical attention. Edge AI helps eliminate the risk with personal protective equipment (PPE). Edge AI is helping to keep energy and utilities workers safe and wearing personal protective equipment.

• Healthcare and first responders: AI and machine learning are helping firefighters, police officers and EMS medics respond to emergencies. Johns Hopkins is even testing a small robot attached to a touchscreen ventilator so people don’t need to wear protective equipment or risk infection when entering an ICU room. National Taiwan University Hospital (NTUH) is reducing the spread of infectious disease with medical panel PCs.

Just the beginning

The challenges we see with edge computing are primarily around dealing with power, timing and environmental constraints to get the right, high-quality data. With any AIoT, automation or autonomous project, it’s important to focus on a business outcome and use case. While this seems obvious, we find that often this initial step can be challenging for our customers to figure out- where is edge going to yield the biggest return? This step is the most critical, otherwise, the project is likely to fail.

Although we have several examples of how edge computing is providing businesses and organizations with value, the industry is just beginning to realize edge computing’s potential and innovate to create faster, more reliable, and cost-effective systems.

Are you ready to explore the possibilities of leveraging edge in your operations?