What edge computing means for business IoT

Limelight Networks VP explains why edge computing isn’t enough to make IoT meet business requirements

This is a contributed piece by Nigel Burmeister, VP at Limelight Networks

Over the last year, the term ‘digital transformation’ has become a popular buzzword. With so much hype around digital change, it’s surprising that a recent global survey of IT professionals revealed only eight percent of enterprises consider themselves truly digitally transformed and 23 percent are still in the early stages.

This means 2018 will be another year of digital transformations as more businesses execute technology shifts and better understand how innovative technologies, such as the Internet of Things (IoT), complement current operations. However, IoT use cases often drive decentralized, distributed computing that’s performed closer to the connected devices themselves. In these scenarios processing is often delivered “at the edge” of a network rather than in traditional data centers or the cloud, which take longer to compute and send critical information.

Edge computing is certainly an important component of IoT, ensuring greater efficiency in processing data – but it isn’t necessarily enough to make IoT meet business requirements.

CIOs are now used to cloud computing, but here comes edge computing. Should they worry? Check out: Edge computing 101: A CIO demystification guide

The pitfalls of relying on the public internet

IoT depends on the fast, secure movement of critical information – and relying on crowded public networks is inefficient and less secure.

Impact on performance / latency

Since the public internet is a shared resource, it’s susceptible to congestion. While we think of the internet as a giant, inexhaustible virtual resource, the reality is that internet capacity is finite. With more connected devices and growing appetites for more complex content, consumers place more strain on this capacity, causing network congestion that significantly slows things down.

Take, for example, online video. Cisco projects video will make up 82 percent of consumer internet traffic by 2021. Video content alone already consumes a large share of internet capacity today and will only increase as more streaming capabilities are added and adopted such as video surveillance, ultra-high-resolution formats, virtual reality and augmented reality. All of this content congests the public internet, and just as more highway traffic leads to longer commute times, congestion on the internet leads to longer wait times (latency).

For enterprises deploying IoT on public networks, latency can lead to serious issues. IoT use in the enterprise often requires low latency, such as an application that’s monitoring a gas pipeline for leakages or a remote device monitoring a patient’s clinical condition. In these situations, delays could be deadly.

Enterprise firms jostle for share of burgeoning IoT space as computing moves to the edge… but What does $1 billion buy you?

Security implications

While public internet use can cause serious latency issues for IoT functions, it’s also important to consider the security issues to which the public internet exposes businesses. The pervasive use of connected digital technologies across businesses has dramatically increased the potential of malicious attacks. In particular, IoT presents new security challenges as numerous devices with minimal security measures constantly connect to the internet.

Hacks that have taken place into connected cars, monitors, cameras, thermostats and more show the vulnerabilities businesses expose themselves to when deploying IoT devices on the public internet. Think of it this way: the public internet was designed for open communications, and this type of infrastructure makes it inherently more vulnerable to cyberattacks and hacking.

 

Bypass the internet and take crucial data to the “fast lane”

To improve the efficiency of IoT-enabled communications, businesses should start by implementing a private network that bypasses public internet congestion and security vulnerabilities, and puts data in the “fast lane”. Leveraging a private network will allow businesses to create a strong framework for IoT and edge compute initiatives.

To supplement the private network, businesses will need to consider additional features like edge presence, or having points of presence (PoPs) globally, as well as capacity and connectivity. Many enterprises are also beginning to discover how content delivery networks (CDNs) can help support efficient edge computing. CDN workloads are primarily egress, meaning that the data is sourced from few origins, but delivered to many endpoints. Meanwhile, many edge compute and IoT workloads are primarily ingress, in which data is collected from many endpoints. This natural synergy allows the data processed in edge compute workloads to be immediately hosted, getting rid of the need to develop additional communication infrastructures. All of these elements create a strong support system, while the private framework isolates traffic flows away from the congested and vulnerable public internet.

Take, for example, Industrial IoT (IIoT) applications of edge computing. For organizations managing industrial infrastructures – such as the gas pipeline previously mentioned – IoT devices can help identify potential issues faster than ever. This technology must have the infrastructure in place to quickly process massive volumes of video and image data to quickly detect and share alerts on leakages or other anomalies. Edge computing coupled with private network connectivity provides a faster, more secure means for IoT devices to process this kind of vital information.

 

Private networks will help make IoT a reality in 2018

Digital transformation depends on innovative technologies such as IoT, but organizations will need to implement the right enabling technologies to effectively support next-gen infrastructures in 2018 and beyond. IoT devices may require low latency and high security, which the public internet simply can’t provide. When businesses need to ingest and analyze masses of data at the edge they should ensure they have a strong supporting network infrastructure that provides low latency and robust security to efficiently manage and deliver this critical information.