Internet of Business says
The global Internet of Things (IoT) market will be worth $1.1 trillion by 2025, according to data from GSMA Intelligence, and technology vendors are keen to get their slice of that pie.
HPE is certainly one of them: it has been building up its IoT portfolio over the last few years via acquisitions, and announcing new partnerships in the sector too, including deals with ABB, GE, and PTC.
Internet of Business caught up with Colin l’Anson, HPE fellow, to discuss the company’s IoT strategy.
Internet of Business: Tell us more about your role…
Colin l’Anson: “I’m one of HPE’s fellows and leaders of the technology part of the business. I help to explain to others what we do and decode the deeper technical information about outcomes for our different customers.”
What is your involvement with the IoT?
“I got involved with the IoT about three years ago and at that time it was a seriously hyped term. My job was to answer whether this was going to be a valuable business for us, and look into what it actually was – what we needed to learn to get into the business, and how we should go to market.
“We had to look forward at what the IoT is going to become rather than focusing on what it was. Three years ago it was still in its infancy, but now there are real projects which we will see results from – although this is still the beginning, the growth stage. It’s the ‘teenage years’, during which some projects are really going to cause us trouble!”
What do you have to consider as a supplier when it comes to the IoT?
“One of the things we have to be wary of is that almost everyone we touch and see has a great IoT idea, but it is just one of thousands. What you’ve got to really get to grips with from a supplier perspective is: Where is the money?, How is this going to work?, and, What is going to be special about this?”
So what has HPE got in its portfolio to help answer those questions?
“If we take it from the edge and work in, at the very edge we have Aruba [which HPE acquired for $2.7 billion], which provides Wi-Fi products, and we have a lot of IP and knowhow to speed the admission of data into the network, so that we can start processing it in the simplest way possible.
“For example, if you’re putting 10,000 devices in a factory, there needs to be a way to register and identify each device, because the data coming from all 10,000 at once is not going to be manageable.
“You have to be able to interface cameras and IoT devices into the network, and we have a gateway and AI product that allows us to do this without people having to sit at keyboards.
“So we admit the devices we know because of their signatures, and the software can distinguish that it’s a camera, for example, and then it checks if the device is using the right protocols. Then we can say for certain it’s a digital camera and we don’t need to raise any further questions.
“We can put this into all types of networks, such as campuses, airports, and other companies.”
What do you do with all that data?
“The flood of data you get is so large that you have to start processing it, and of course we have edge processing, which looks different to what you would see inside a traditional data centre.
“The physical structure and boxes look similar and they are built to exist inside industrial or office buildings, but they don’t need data centres to live in and they are low-power devices.
“We also have a very high compute density inside them and we are capable of running SAP HANA inside it. This means that a manufacturing company, for example, could process a lot of its data at the edge, and so the data volume required is only to do with the insights, and not every piece of data that the company generates.”
How do you work out which IoT ideas are right or workable?
“This is all about getting into a dialogue with the customer. The first thing we need to do is work out what the outcome is for the business, and what they want to achieve, then we can start to understand what technologies are appropriate.
“We can also then understand the conversation about the volume of data, and how complex the computation is going to be. Is this going to be a few virtual machines set in a server in the company’s data centre? Or is this a heavy-grade compute that they’re building towards?
“We’ll calculate the scale of the project by looking at the size of the data sets and how many calculations are required, and then we can work out the scale of the compute necessary.
“After this, we will look at the choice of ISVs that the customer has, and then we can create a pathway with the solutions necessary to solve the business predicament.
“We start with business outcomes, not with people’s ideas.”
Why are some sectors – like financial services – going to be less involved with the IoT?
“The financial services sector is undergoing digitisation – as are many other industries. However, the difference is that currently there are not many data points in the IoT that they can constructively use.
“They’re digitising a big core and the IoT area is small. If you compare it to manufacturing, that sector is also digitising its core, but it has loads of physical touch points.
“Financial services is concentrating on digitising the core to move companies from their very locked systems into a much more agile environment, and as this happens they will be able to bring in much more data, but at the moment they’re just that one step behind.”