The era of the Internet of Things (IoT) is upon us. The long-heralded promise of ‘machines talking to machines’ is already here.The growing availability of cheap sensors, ubiquity of cheap, reliable communications and the ability to process ever greater amounts of data is promising to revolutionise the way we live and work.
Devices such as smart lighting, thermostats, CCTVs and even toothbrushes have all hit the headlines in the recent past, highlighting the ever increasing number of connected devices coming to market.
Looking beyond the hype and the headlines, we can see that industry, rather than the home, may be the biggest beneficiary. The much anticipated fourth industrial revolution, driven by AI, IoT, robotics and cloud computing, is already having an impact on many industries.
For some sectors however, the rapid pace of evolution can be problematic. DIGIT spoke to Steve Aitken, the founder of Intelligent Plant, about his background in software development, the industrial app store, IoT and the oil and gas sector’s approach to the new Internet of Things.
DIGIT: Can you give us a little more information on Intelligent Plant and what it does?
Steve Aitken: Intelligent Plant has been in operation for around 12 years now and we’ve been in the industry 4.0 space from the beginning. We started off building things for clients, so we were effectively ‘guys for hire’. We did a lot of work with oil companies focused on technologies which were their own. So they owned the IP and we built things for them which would help them monitor oil and gas facilities, corrosion of key equipment, all that kind of stuff.
There are a couple of common approaches to software within the industry 4.0 space. Some companies build displays for pre-built tools, add some calibration but effectively just tie existing technologies together. Intelligent Plant took a different approach, we created software that solved a particular problem.
In the early days we built products which would allow engineers to build displays, perform calculations and create visualisations. If I’m being honest, it was not a huge success. We built the software, but not many people bought it. So we started looking at a different way of working. Around the time the oil price dipped, we looked at the market and said, ‘Hmmm!’
We didn’t believe that businesses would continue to pay for individual, bespoke development for much longer. We recognised that it’s actually pretty hard for newcomers to get into the industry 4.0 space. So we created the industrial app store in an attempt to help with that. It provides a cloud platform that enables the sort of development capability we’ve been speaking about, but also gives access to example data sets. So developers can not only build things, but actually test them using real data sets before going out to (oil and gas) operators and potential clients.
DIGIT: The industrial app store sounds like a major advance. Is it something entirely new?
Steve: The industrial app store is a bit different to many other systems that some people can, on the surface, think of as exactly the same.
In the oil and gas sector, we often get compared to GE’s Predix, or Siemen’s Mindsphere. However one of the key differences with the app store is that your data stays on site. Most oil companies already have a historian – a place to collect and collate all of the data generated. Once that data is available, it can be used in the cloud and across any number of apps. Rather than having your app in the app store, you can simply link your app to the app store and the data is stored elsewhere.
So the app is hosted in one place, the app store in hosted in one place and the data is on site. They all communicate via the cloud.
The fact that the app store exists and can be accessed is a major reason we’ve got into a lot of new clients. Many people in the oil and gas sector have become a little bit desensitized to over-zealous sales pitches. It can be hard to tell what’s real, what actually exists. So being able to go into the app store and browse through live apps and see real data gives them a lot of confidence.
One interesting thing is that the growth in users is consistent. The app store has been live for a year and a half, or two years now. I expected the number of users to grow exponentially, to see that curve in terms of new users. However, whenever I check it’s more or less a straight line gradient.
I think that reflects the big challenge for industry in terms of getting people comfortable with connecting and rolling out live plant data to the Internet.
The big advantage of the app store is that it’s real. The disadvantage is that a lot of people within industry – especially oil and gas – don’t really want to put their data on the Internet.
DIGIT: From the outside you would imagine that the oil and gas sector would be pioneering in terms of technology and pushing forward new capabilities to stay at the forefront of production, open up new opportunities and find efficiencies. Is that not the case?
Steve: When it comes to technology, there are a few identifiable leaders within the oil and gas sector, but there are an awful lot of followers. I think much of this comes from the levels of investment involved. When you look at what they’re doing in oil and gas, we’re talking multi-million or billion pound projects, which involve an awful lot of hardware. That breeds a certain mindset which is very planning-oriented and very risk averse.
That same mindset can often been seen on the technology side of things. Few people are willing to be the pioneer and run the risk of failure. What we tend to find is that if a company can see a competitor, colleague or peer doing or using something new, then it’s probably OK to do the same.
DIGIT: How do you combat that mindset? Are things starting to change?
The oil and gas sector has some unique problems, but far more of them are common to all industrial facilities.
We’ve started to see a lot more take-up for the app store in the last six or nine months. When we originally launched,we thought we’d get an immediate response, but that didn’t really happen. Gradually we started to get invited to events, then the operators started to take notice and realise that something new was on offer. It’s taken a good six months or so to get to the point where income started coming through the app store. From there, it’s taken around 18 months in total to transition from a service-based company where we were building software for other people, to the majority of our income is from the store.
So yes, things are starting to change, but in a fairly cautious manner.
It does help that we now have clients that we can actually talk about. We’re doing a lot of work with BP. They’re connecting to the industrial app store and they’re looking to utilise those services around the world.
DIGIT: Do you see this interest in the app store as part of a growing understanding of industry 4.0?
Steve: Our tactic for rolling out IoT and industry 4.0 is to put apps on the app store that address some very industry specific problems. For example there’s a great well intelligence app for optimising well production – straight from the data.
Normally within the industry, companies have all these engineering models, five or six different ones. They run them all and they all depend upon each other, there’s an awful lot going on and to be honest it can become something of a random number generator.
Our approach is entirely different. Our app looks at the actual sensor values, the flow data and applies a neural network to it. It can make recommendations which say that if you make certain adjustments, you’ll get this much more production.
There are two major advantages to that approach. First it can provide real insight into the process and highlight things that the company does not know. It also provides immediate advice on how to improve the situation. For a lot of people, that’s entirely new. It’s a perfect example of how IoT and industry 4.0 can use sensor data in a real time way to make things better.
DIGIT: So it can be the company itself which stands in the way of industry 4.0 and IoT adoption?
Steve: Yeah. It’s easy to lose sight of the fact that these things on their own don’t necessarily make a difference. Sometimes even a buyer doesn’t realise that.
Some companies might buy a tool system that’s going to give them real-time data and provide useful insight and advice, but if they don’t pay attention to the device then nothing’s going to happen.
There’s still a lot of work which needs to be done around planning and making sure that the organisation has the structure in place to enable them to make changes based on the information they get out. In certain situations you have to make sure they actually have permission to act upon the data they get out and make changes.
In some cases, getting that permission from the company, to act on insights can actually be one of the hardest things to do. It can be incredibly frustrating. If you have permission to do something and nothing happens, then fair enough. You can say it simply didn’t work. However, if you run something, but you don’t have permission to act on the output and actually make changes – then you’ll never know if it would have worked or not. It can be incredibly infuriating and frustrating.
DIGIT: How do you see that situation changing?
Steve: There are two sides to the way this whole situation is evolving, the first is marketing. There’s a lot of hype around IoT and in some cases the term can be misused, in the same way the term ‘big data’ can be applied to situations where are not actually big data at all. At a basic level, when I think IoT, I think about cheap sensors, cheap hardware that communicates back to some kind of system. When you look at it in those terms, I see a lot of actual IoT in the home and and hobbyist environment and not much of it in the industrial environment at all.
DIGIT: Really? That seems at odds with the reports in the media
Steve: When you look at this with regard to offshore oil and gas facilities, its a hazardous area, its much harder to get hardware on there.
As an example, the industrial app store supporst MQTT [the IoT connectivity protocol] . You can have have devices post data to a secure locker, then use that data to build apps of all descriptions. But there are zero oil and gas companies that even interested in that.
One of the biggest challenges the oil and gas sector has at the moment is the use of data. Companies have been recording IoT style data from sensors and systems for maybe 15–20 years. In a lot of cases they haven’t done nearly as much with it as they could have with it.
The value of having such an incredible data set can often be missed or overlooked. It’s only when we come along and ask , ‘Have you got the last ten years?’ that we find out they’ve actually deleted it.
It’s incredible. It’s like deleting research. If you have access to that data for alarms, events, control systems and the like, then you can do some very clever things. For example, you can pick a specific alarm and then go back through the data and see what tends to happen before the alarm activates. You can use that to draw a cause and effect chart and figure out probable causes.
What’s cool about that is, if you have loads of historical data, you can use that to look at something that’s happened multiple times and say here are all of the things that caused it.
But if you don’t have that historical data, then you simply can’t do that. In a lot of cases, companies are not seeing the value in the data because they’re for something incredibly basic, like counting the number of alarms. They use the data to find out how many alarms have there been. That’s it. Job done. Outside that there’s no value in it.
We’ve seen similar situations time and again. And it’s not that storage is any great problem, or is costing anything. Instead, the issue tends to be that the tools that are being used to store these alarms and events are in an older style where the indexing is not that great. So the larger the data set, the slower it gets. So some companies tend to delete the older data because allows them to access the new data more quickly.
So this is an area where industry – oil and gas and others – can learn from the Internet of Things, in terms of dealing with massive amounts of data, using new NOSQL databases and things like that.
There’s an interesting contrast between the evolution of IoT and the way certain industries work. On the IoT side of things, we have the ability to store incredibly large amounts of data, thanks in part to the fact that IoT was introduced in a period when storage was very cheap. However, on the industrial control systems side of things, the intelligent thing to do was to to filter the data, so you only stored the things you really needed to, to be able to get the same data out on the other side.
It’s quite interesting when you combine those two. When you combine the IoT efficient storage thing, with the industry filtering and understanding of what the data actually means, you get something very powerful.
To be honest, most of the unique insights are happening within individual teams and departments at the moment. However, you don’t really have to look very far to find a solution which someone else didn’t even know was a problem, when you have access to this sort of oil and gas data.
Getting access to the data can a very big deal though. Some operators we’ve spoken to for around two to three years before we can get anywhere near their data. In one case, we were given access to a company’s historical data and were generating reports for them using that data, but we were not allowed to share that data with other people, other teams within the same company.
DIGIT: That seems challenging. Are you seeing any change in the industry? Who’s pioneering?
Steve: The Oil and Gas Technology Centre (OGTC) is making a real difference here. They have much of the industry involved, backed by significant funding. This gives them the ability to share activities between operators, identify common problems and then help to find – and fund – solutions. That’s something new, something no one else has been able to do, because the collective approach can really take the sting out if a project if it fails. That’s one of the reasons very few people are willing to pioneer and try new things. You can be putting your career on the line.
The OGTC is helping a lot in this space. A lot of their purpose in this sector is to move the industry forward in its thinking, to the point where it’s not saying ‘why should we share this data?’, and talking about reasons why it data should not be shared, but instead looking at what data can actually be shared and what can be done with it.
The problem in the oil and gas sector is that the focus tends to be the wrong way round. In companies such as Google, the IT department, the development departments will all be more or less aligned. There will be clear set of priorities which will go right across the whole company. That’s not the case when it comes to oil and gas.
If you look at an energy company, they’ve got clashing priorities. The engineering job will be to produce more oil, minimise downtime, reduce risk to the plant, people and environment. The priorities for the IT team will be to create a safe environment, make sure everything works, everything’s secure. So in many cases that means locking the data down, making sure it doesn’t move.
The IT side of things can have absolutely no idea of the full benefit to the business which could be gained by changing things and doing something more innovative. So they can’t weigh up the benefits with the possible downside. All they can do is focus on the downside and say ‘it would be really bad if this happened.’ So it’s seen purely as a risk. It never happens.
But the probability of the negative outcome are only so high. If you look at the possible gain, multiplied by the chance, then it’s clear from a business perspective that the company would be far better off by trying something new. That tends not to happen because you need to go so far up the organisation, that the decision has to be made at a director level and that’s quite hard.
Which is why we end up speaking to them for about two years…
DIGIT: More and more companies, organisations and industries are now saying that the future is all about collaboration. Are you finding that’s the case?
Steve: I was speaking to one operator recently. They were keen to take one of the services from the app store – which is online and available to anyone and which you can start using within 10 minutes – and they wanted to turn it into something which they would install locally.
It can be quite eye-opening. We’ve spent a lot of time solving problems and making sure that we can get data efficiently and securely from one side to the other and some companies would much prefer if it didn’t do that and just acted like with was the 1970s.
One of the interesting things which came out of that discussion was that: ‘it might be better for the industry to have the ability to do these things swiftly but looking at ourselves it might be easier if we do it the hard way.’
It can be quite a challenge.
Where we are starting to see a real difference is where there’s someone within the organisation’s IT team – at a fairly senior level – who has a background in engineering .
When that’s the case, amazing things can happen. Those people are the ones with the perspective to say: ‘it’s worth taking the risk, because I can see, right down the line, exactly what this is going to be be worth. And I know that even if there is some kind of issue, it’s not going to be a big deal.’
In those cases we can see some brilliant things happening, but when senior IT decision makers come from a pure IT background, then it can be very hard.
DIGIT: What’s Intelligent Plant doing to change this?
Steve: In the good times, before the oil price went down, one of our major limiting factors was access to good people.
We set up a partnership with Aberdeen university to help address that and we got involved with teaching. We set up the Innovation Prize to encourage students. We even set up work placements to help that talent pipeline.
A lot of this is directly relevant to realtime data, what I now term IoT data, because it’s very similar. We’ve maintained our relationship with the university and built upon that. The university is now running an AI course which will also be hugely important to the IoT sector as well.
Steve will be one of the speakers at DIGIT’s IoT Scotland event, which takes place in central Edinburgh on May 30 2018.