This has become known via the buzzword ‘Internet-of-Things’. And whilst M2M analytics has been around for decades, the IoT allows all sorts of companies to do this on a large scale. The most notable difference between M2M and IoT, in my mind, is the ability to create new services for customers by connecting up products that traditionally wouldn’t be feeding data via the Internet.
This market is growing in importance and vendors are increasingly making a play for $ spend with customers - as we have seen via a number of acquisitions and new product announcements from the big players.
As such, it was with interest that I saw 451 Research’s latest report on the Internet-of-Things outlining a number of the trends and challenges facing companies implementing IoT projects.
On the bright side, the report, which surveyed nearly 1,000 enterprise IT buyers worldwide, found that 71% are gathering IoT data today - up three percent from the previous quarter. In addition to this, those already with IoT initiatives underway expect their mean IoT related spending to grow by 33% over the next 12 months.
The research found that 90% of enterprises will increase IoT spending over the next year, and 40% of buyers will raise their investments by 25% to 50%, when compared to 2016.
However, as with all new technology, it’s not all plain sailing for buyers. IoT projects have the potential to get very complicated, very quickly, because of the masses of data that can be collected in a very short amount of time. Equally, companies are often unsure where to apply their spending, as IoT can have a number of diverse applications.
Throw into the mix that a cultural change is required by companies looking to shift from products to services, and it’s unsurprising that enterprises have some concerns.
The number one worry identified by 451 is that of security, with 50% of respondents citing it as the top impediment to IoT deployments. With the ability to connect up literally anything, companies are very quickly broadening their risk. With more devices connected, more potential threats follow suit.
However, of more interest is that a significant 41% said that IoT’s lack of perceived Return on Investment (RoI) was a barrier to adoption.
I spoke with Ian Hughes, IoT analyst at 451, about why this is. Hughes explained that companies that are incrementally experimenting with IoT are likely to be more comfortable with the RoI question. Whereas those that invest in it to solve big problems, or to ‘digitally transform’, will be more worried about the returns. He said:
I think when a company looks at it as a new thing that they have to install and implement and roll out, then it becomes a different discussion to when its an incremental rollout of IoT. To say ‘here’s a new thing, it’s going to cost some money’, they then challenge why they want to do that. Whereas most of the IoT stuff that we see is this incremental approach, so things are gradually coming in.
IoT also often bounces around industry verticals, which often confuses people. It’s power is being able to pull things from more than one place and decide what to do, not just zone in on a thing and that’s a tricky thing for people to navigate. That’s why a lot of it comes through via data centres, via buildings and via industrial processes.
We’ve got lots of buzzwords - we’ve got cloud and we’ve got big data - well, IoT feeds big data. And big data doesn’t care where the data comes from. So when you apply analytics and machine learning to that, that mass of data is when you start to get some benefit. You notice that one particular part of your business that would appear identical to another is actually performing better.
You realise that some of the peripheral things that you weren’t looking at before you can tune to make other parts of the business better. It’s that wide disparity of data is what helps IoT - but that’s why it’s difficult to say, right I’m going to install an IoT system. It might be something outside your normal kind of business that can tell you that stuff.
With regards to where spend is happening, 451 said that IoT deployments and usage is particularly strong in enterprise initiatives around data and transactional intensive workload categories, such as data analytics and security. It is also seeing IoT-specific projects around the data collection and analysis of financial, healthcare or industrial functions; the uptime/reliability of mission-critical line of business servers and applications; as well as monitoring the efficiency and costs related to a specific business operation or department such as a hospital emergency room.
Some 68% of companies said that they are currently using IoT data to optimise operations, such as improving performance, or driving efficiencies. This is often the first stage of diving into IoT. Whereas, 42% of enterprises use IoT data to develop new products or enhance existing products and services, according to 451 Research. This group will likely be more sophisticated in their approach, in my opinion.
However, one of the most interesting stats to come out of the research was with regards to skills - which seemed to divide opinion. 54% of respondents said that a lack of trained IoT staff is not an issue for their organisations, versus 46% who said that they are having difficult filling IoT-related positions.
Hughes said that this latter group are likely to be companies that are trying to shift from products to services through the use of IoT. He said:
That is an unusual one. The IoT skills, I suspect that some people don’t know what skills they haven’t got.
I suspect that those that aren’t struggling as much are deeper into it. They’re the core IT people. It’s when you’re in another industry, you’re a tyre manufacturer or something and you want to make your tyres become a service, that’s a whole slew of things that you maybe haven’t been involved with.
Going from a product to a service is quite a leap. It’s not just the IoT side of things, it’s everything on that digital transformation side of things. It’s a bigger leap. Whereas the software or IT industry has always been a service, and now it’s providing more services.
Finally, Hughes also wanted to highlight that one of the biggest challenges going forward for companies will be around the management of their IoT data and projects. He believes this to be the case as more enterprises realise that they need to manage data both at the edge, and centrally in the cloud. Hughes explained:
One of the biggest challenges is that there has been a general assumption that you take all your IoT data and you put it in the cloud. So you bring everything to a centralised point. And there’s not really enough infrastructure to deal with the amount of data that can be generated from all these pieces. So it doesn’t make sense to do everything centralised.
So we are back to trying to distribute that compute power and distribute that intelligence and distribute the analytics, so that it is closer to the edge sometimes. That’s quite tricky to build and to manage when people are just getting used to centrally managing things. Now this is about device management, gateway management, distributing your business rules across a wider area.
Particularly if it’s technical data, you need it close to where the things are happening, you don’t want any latency to and from the device. If you take a driverless car, you don’t want it to apply the breaks by asking a server first. You need it to respond close. It’s not one or the other, it’s both. And that management is tricky.
As I’ve said before, the biggest challenge for companies as it relates to IoT - aside from the technical/security hurdles - is that shifting from a product orientated business to one that sells services is not easy. It requires different thinking, different sales approaches, different CRM and a different approach to manufacturing. That should not be underestimated.