Q+A: Hitachi Vantara on the changing IoT, big data and AI landscape

In conversation with Hitachi Vantara experts Jonathan Bowl and Wael Elrifai at Big Data World and Smart IoT 2019

IoT, analytics and AI are not exactly new technologies. But now that companies have incorporated them in to their business processes, to various degrees, why are so many failing to achieve the outcomes they expected? Last week, at Big Data World and Smart IoT London, Techerati editor-in-chief Alice MacGregor and second in command James Orme stole two of Hitachi Vantara’s keynote speakers, Wael Elrifai and Jonathan Bowl to talk business outcomes, the changing tech ecosystem and the enterprise cost of failing to leverage data to its full potential.

James: What are the key trends you’re observing in the industry at the moment and what have been your key takeaways from the show?

Wael: The first one I noticed is that there’s starting to be a shift away from technology more to outcomes. A lot of the companies in the past, and go back about three years, you attend some of the big data conferences, there was an overwhelming focus on Spark or Flink or MapReduce. Things of that nature, technical things. Now there really is starting to be a focus on what is the outcome and then afterwards the how.

Alice: As regards the ecosystem, have you noticed a different exhibition vibe on the floor, are new companies emerging?

Wael: One of the big things we’ve noticed in the industry, and it’s reflected of course in the show today, is that there’s been a lot of consolidation; acquisitions and mergers. These aer indicative of the shift away from tech towards the solution.

Jonathan: What I’ve noticed from a lot of the presentations yesterday in the room that I was in, is that people are still grappling with the same challenges and people are not moving as fast as everyone would like. I suspect, although I wasn’t here last year, that people are still wrestling with the same challenges as in 2018. When I heard various customers talk yesterday it’s clear they’re not getting the value from the data that they want to. They’re not getting the insights.

People keep talking about data being data-driven, but the reality seems to be a lot of organisations are still stuck in the monitoring phase. Understanding what’s happening in their business rather than actually using the data analytics to truly power their business, if that makes sense.

To Wael’s point, there does seem to be a lot of movement in the market at the moment, which is quite interesting.

Alice: Do you have any thoughts on why that might be, the changing dynamics there?

Jonathan: It probably plays to what Wael is saying. There is a lot of change in the industry as well as consolidation. If you were here a year ago, it would have been inconceivable in that era that Hortonworks and Cloudera would merge and IBM would acquire RedHat. No one would have guessed that. You’re seeing so many spinoffs at the moment from the organisations as well. It’s a very hot and competitive in the market at the moment.

Alice: If we look now at Hitachi’s role in the industry, in this space, and how you’re leading a new IoT revolution, can you explain a bit about that?

Wael: Hitachi is at a very special place in this industry. We mentioned earlier companies like Cloudera and Hortonworks. These are technology companies, these are data companies, these are computer companies. Hitachi is all of those and a manufacturing company: a 300,000-person, $85 billion a-year manufacturing and technology behemoth. We are pretty specially placed in this convergence of information technologies and operational technologies, IT and OT. We’re very specially placed in this market.

Alice: Would you have anything to add to that Jonathan?

Jonathan: It’s very relevant and we use it so often that it’s just lost in the noise but you are combining these 50 odd years of technology expertise – genuinely, that is what it is – with more than a hundred years of operational technology expertise. What I’ve seen, and I’ve been at the company now for up to three years, is that you can see it’s really starting to manifest itself in the projects that we are working upon.

We are working with the other divisions of Hitachi and using all of its skills and knowledge. We have software, we have a platform, we also use data scientists and data engineers but there’s all these other businesses there with expertise around water, manufacturing, energy, transportation, an supply chain that we can prevail upon to support the projects we’re working on with our customers.

Invariably, our starting point is to try and get a problem for our customer. To quote an exam question, what is it we’re trying to fix? We look at the assets we have together with the wider Hitachi group and see how we can combine those assets to try and address the problems of our customers.

Wael: I’ll add one thing to that. This is reflected in the fact I came from a company called Pentaho who were acquired by Hitachi about three and a half years ago. I remember the business that we were in three and a half years ago, which was a very tech-focused business. Just a pure tech business. Data integration, big data, that was it. Now, to what Jonathan was saying a moment ago, we are focused on solutions, and industry-specific solutions.

Because none of the companies we work with are in the business of just tech. They’re providing some value to their business customers or their consumer customers. Because of the size of Hitachi and the level of knowledge at Hitachi we’re able to reach out to the rest of the organisation and get some of that operational expertise, it really puts us in a very special place.

James: Wael, you’ve written a book about IoT and have spoken a lot about IoT being a societal imperative, a social imperative. In what ways are IoT and AI transforming the lives of the general population?

Wael: AI is a term that’s used a lot now, every company says they use AI in their software now or use AI in their business. The joke I heard was that if it’s on a PowerPoint, it’s probably AI and if it’s in Python it’s probably machine learning.

The first thing I would say is that it absolutely is a societal imperative because ultimately there are efficiencies that come out of using these techniques.

You go back a hundred years and companies were operating on steam; factories were being operated on steam power.

You organised the factory based on the idea that it was being run on steam power, you put the layouts in a certain way. When there was a shift to electricity, they just electrified the plant. They pulled out the boiler. There’s wasn’t a boost in productivity, it didn’t happen. The same thing happened with the introduction of computers in the workplace. There was a change in technology but there wasn’t a common change in business transformation.

One of the things that we’re seeing here and one of the big trends we’ve noticed with ML and AI is that we’re getting to the point where we’re actually seeing the business transformation that’s required to actually leverage the technologies and techniques we have now.

James: And what industries are benefiting the most from this transformation AI in IoT?

Wael: The ones that have been the biggest adopters of machine learning in particular, and when I say that I mean supervised learning, unsupervised clustering and things like that. The biggest adopters have been the insurance industry and it’s easy to see why. Their entire business is statistics and machine learning is more sophisticated statistics.

The biggest beneficiaries long-term are going to be companies that own stuff, which is most companies. Manufacturing companies, car companies, people who are building stuff. Companies that operate in small margins and that need streamlined operations need to be able to predict downtime in their systems, they need to be able to optimise quality assurance and quality control. Those are going to be the ones that see the most value.

Alice: What can those companies do to ensure that they’re really optimising and getting the most value out of these technologies?

Wael: I would start by saying that they should be looking at the parts where they already know the business transformation process.

Let me give an example of the factory I visited last year. It’s a steel rolling mill and what they do is they’re having to press this steel and they’re building these coils up and then putting them through the picking line and the hot section and the cold section, all these different things. Then it was telling me that they have about four percent downtime. That’s catastrophic in the steel industry, which is a tiny margin industry.

How can it predict this downtime? The control room literally starts shaking about 20 seconds before a failure, because the coil has gone down the wrong part of the line.

Those companies they know how to fix it once it’s broken down. In fact, they know the business transformation as far as maintenance goes, because have regularly scheduled maintenance.

What they’re able to do now is apply the data and use techniques so that they can start predicting that “it’s not just the failure right here, it’s not just a screw that’s loose right here where the thing went off the rails, it’s actually being caused by something that happened three days ago.”

James: Jonathan, you spoke in your presentation about what it takes to become a data-driven company. In your opinion, are businesses currently using this data in the right way? Are they enacting the right processes and how are they failing to achieve their goals?

Jonathan: It feels to me that the term data-driven is overused to a certain degree. It’s not just about technology. All this amount of culture shift, there’s got to be another term. It normally starts from the top. The hypothesis almost is, “If you’ve got a chief data officer or digital officer who’s sitting on the board, you have got a board level agenda that’s driving all the initiatives further down below.”

We were saying this morning that traditional organisations have got this wealth of information. It’s almost criminal not to be able to use all that but still there’s so much data, and they call it dark data as it’s not being used.

If you’re a startup today what’s the one thing you they crave more than anything else? It’s data.

They’ve got all these ideas but what they’re craving is data. They’d give anything to have 30 years of call data records that a telco’s got. They’d give a lot to have huge amounts of data on what the retailers are doing and a whole bunch of models around customer buying patterns. They don’t have that but the traditional businesses do.

Someone was mentioning yesterday about data science and how important that is. We ask most organisations, “What’s your budgets for data science?” They might have a graduate here and some others here. Are they taking it seriously enough? I don’t know. It seems to me that the organisations who’ve got a very clear intent, who’ve got someone on the board who’s helping to drive these initiatives are the ones making a bit more progress.

Alice: What are the stakes for boards who aren’t listening to these new demands?

Jonathan: That’s a good question. There’s some stats that I’ve seen over the last couple of days that said 50 percent of the fortune 500 companies from 2000 are no longer in existence since. These questions really should be front and centre and quite why there’s not a sense of urgency from everybody it’s hard to say.

Wael: There probably is a sense of urgency but there’s a lack of knowledge of how to execute. Everyone wants to, we say, roll their own and try to make their own thing from the ground up. That’s typically not a good way to approach anything. I mean at Hitachi, we don’t design our own building that we’re going to work out of, we let the company that’s expertise in engineering and construction do that.

I think companies, whatever business they’re in, if prediction and data science is something that they need, then they should be talking to experts like us about that so they can focus on their business and the business transformations required.

Jonathan: The point is that there is no silver bullet that’s going to fix everything. Just start with a small problem and try and fix that and don’t try and do everything all on your own.