Could the IoT be fragmenting into increasingly disparate categories and terms, asks Adrian Bridgwater?
First there was the Internet. Then, there was the Internet of Things (IoT). Sometime after the IoT had established itself as an army of sensors, devices and equipment with a connection to the web, there arrived the notion of the Industrial Internet of Things (IIoT), which seeks to distinguish the use of IoT technologies in factories and power plants, for example, from the consumer realm of wearables and smart home gadgets.
But just as some industry watchers were starting to think that the IIoT term would serve their classification needs for a while longer, yet another layer of IoT terminology has surfaced: the Industrial Infrastructure Internet of Things (IIIoT).
What is the IIIoT, anyway?
As a notion of a defined space in the stacks of data and applications feeding to and from the IoT itself, the Industrial Infrastructure Internet of Things (IIIoT) is characterized in two ways: by the types of installation in which it is found; and by the nature of the analytics that it performs.
First, by installation type, the IIIoT is the optical cabling in smart cities, the networking layer of energy plants and ‘big’ machinery from printing presses to industrial moulding units. The IIIoT is digital roads (when they finally arrive), digital street-lighting and the digital air conditioning system in your favorite airport. Often representing a part of what might be defined as a civil engineering project, the IIIoT is massive in physical scale and massive in the number of data points it creates.
Second, by data analytics type, the IIIoT is characterized by meta-level data analytics. But why is this so?
Machines produce more data all the time, obviously. What this means today, 17 years into the post-millennial age, is that we are moving from terabytes onward to petabytes. By the end of the decade, we will be discussing the move from petabytes to exabytes.
What this volume of data creates is an opportunity for meta-level data analytics. Industrial infrastructure, once digitized, can start to shoulder some of the analytics we need to use within its own slice of the total data fabric. Applications will then feed from a more tuned, refined, analyzed, intelligently automated (and essentially smaller and more accurate and de-duplicated) pool of data. The IIIoT shoulders the first stage of all data analytics by executing it at the meta-level.
For an opinion on this topic, Internet of Business spoke to two firms that operate in this space. Yotta is an infrastructure asset management company that uses IoT technologies to collect data in towns and cities, working with local councils to maximise connectivity and value from data analysis. Meanwhile, AspenTech uses IoT, AI and big data to optimize and maintain energy plants, allowing large energy firms to identify when machines are most likely to breakdown before they do so.
Thoughts from the edge
As Yotta chief product and technology officer Manish Jethwa explains: “It can be easy for data to become unmanageable when the quantities continue to rise to such a high level in terms of terabytes and petabytes and there must be a sound infrastructure in place to mediate the data, which is where meta-level data analytics at the IIIoT level can help.”
He agrees that businesses can use the cloud model to help reduce the sheer amount of data being gathered by sensors and probes, but it is important to analyze what infrastructure is being used to hold and manage the data.
Yotta’s asset management platform Alloy works to extract large volumes of data collected through the use of microservices. The use of microservices is critical because it can help to filter smaller elements of data and then drive crucial data to the right places, which then allows data analysis to happen at a more general level.
“One simple example of this type of data collection in practice is when collecting temperature variations within cities, which may require a number of different sensors to collect regular readings. Microservices can provide a valuable service in reducing multiple measurements into key notifications of predefined threshold being exceeded. It is easy to imagine similar technology being used to monitor noise and air pollution too,” said Jethwa.
Mike Brooks is senior business consultant at AspenTech. Brookes agrees that by using the vast volumes of data today in the right way, businesses can now arm staff with the intelligence that pinpoints exactly the specific part in an asset or system, that if worked on today, could help you to avoid unnecessary failures during a spike in demand, or enable you to negotiate the risk of a random shut down.
“The science of maintenance within the Industrial Infrastructure Internet of Things leverages both historical and real-time operational data, that when fed to algorithms, can model the precursors to failure across all assets and systems. This enables truly accurate and proactive identification of asset vulnerabilities in near real time. The output is a refined set of recommendations that enable engineers and maintenance professionals to act well ahead of any potential impact on individual assets or larger systems,” said Brooks.
Could we add even further I’s to the IoT? Hopefully not… although we’re not ruling out the likelihood that certain industry players might not be able to resist the temptation to start talking about the Industrial Infrastructure Internet Intelligent of Things (IIIIoT).