The multifaceted utility of the Intelligent Internet of Things

Max Smolaks

August 7, 2019

5 Min Read

How Shell uses analytics and AI to prevent waste and fraud

by Jelani Harper 7 August 2019

Manifestations of Artificial Intelligence have long been predicted to revolutionize the Internet of Things, culminating in something called the Intelligent Internet of Things (IIoT) that maximizes business value for machine-to-machine communications.

Actually seeing the impact of the IIoT in contemporary use cases only confirms this assumption.

According to Cybera Chief Marketing Officer Bethany Allee, “To remain competitive, retailers are now required to integrate IoT into their storefronts. And that’s everything from [gas] pumps…and things like outdoor payments, to also things like network-enabled coffee pots.”

Cloud analytics

The combination of different aspects of AI with the secure network connectivity required for taking full advantage of the IoT can deliver benefits like better use of cloud computing, heightened regulatory compliance capabilities, and decreased incidence of fraud - examples of which are detailed below. In many of these instances “things have some element of expert systems and machine learning in their apps,” added Cybera President Cliff Duffey.

Cloud computing is the foundation of the benefits delivered by the IIoT. The connectivity required for secure deployments of cloud-based AI applications is critical from both the safety and consistency standpoints. Moreover, this connectivity has become as diverse as the IoT devices themselves. Duffey mentioned that in the oil and gas industry, for example, “those devices used to have real simple connectivity, just to some in-store systems. Now, they have to be connected up to the Internet as well.”

The intricacy of this connectivity is exemplified in a Shell use case in which the network "is supporting all of the new devices that need to be connected to the Internet, like the fuel dispensers," Duffey said. “We also have to do the network connectivity for the dispensers to the in-store point of sale and then several Internet destinations.”

After moving the relevant data from such deployments to the cloud, organizations can perform cognitive computing analytics. In the case of Shell, the company was “tracking and monitoring fuel data, as well as cartels and so on, unique to fuel,” Duffey said. “One of the things that a couple of different cloud companies do is they’ll gather the data on how much fuel has been sold and then what are the fuel levels in the ground. If there’s a delta of 50 gallons, then something happened. It didn’t just disappear or evaporate.”

Regulatory realities

In this use case, deployment of cloud-based cognitive analytics is instrumental for regulatory compliance. According to Duffey, it can help draw conclusions like "maybe there’s an underground leak in a tank, which for EPA and environmental reasons is a really big problem.” The reality is that companies must monitor such data — and leverage it to prevent any wasteful incidents.

In the oil and gas industry, “a lot of the states in the US require that there be daily or weekly checking of the fuel levels to determine whether there’s an environmental leak,” Duffey explained. Machine learning and other analytics methods applied to IIoT data underpin “cloud applications that can monitor in real-time, to notice if there’s a potential leak quicker.” Moreover, the Shell use case involves interconnecting IIoT devices with the Europay, Mastercard, and Visa (EMV) network, which enables the company to expedite adherence to 2020’s outdoor EMV chip card compliance deadline for its fuel pumps.

Fraud prevention

Another defining feature of the IIoT is the robust, reliable connectivity required to feed intelligent algorithms to the cloud. The Shell use case typifies this fact: the aforementioned outdoor EMV network supports chip payments at fuel dispensers, which is still somewhat novel within the industry. Chip payments were designed to prevent fraudulent credit card transactions. Whereas the magnetic strips of these cards are easily copied for conventional credit card fraud (which is well-suited for use at automated fuel pumps), it’s much more difficult to copy chip data — which inherently reduces fraudulent activity.

“If you’re going to read the chip based information, you have to be able to communicate that immediately,” Duffey explained. “If the network is down, the dispenser simply can’t pump fuel anymore.” Developments in wireless communication such as 5G or even 4G connectivity enable such use cases, so that “even if the primary Internet connection goes down, we can send the chip based payment over a wireless connection through a partner like Verizon.”

Here to stay

The Shell use case and other examples throughout the oil and gas industry prove the IIoT’s potential in the enterprise. It offers means of leveraging cloud-based analytics to produce business value, ensure regulatory compliance, and prevent common forms of risk, like fraud. It requires ongoing connectivity — and a few backup plans to stay online.

Keep up with the ever-evolving AI landscape
Unlock exclusive AI content by subscribing to our newsletter!!

You May Also Like