Brad
Identifying the customer’s pain point or business objective is the beginning of the journey map for digital transformation. With this defined, we can position the journey’s start appropriately. Hitachi Vantara’s role is to unlock the synergy between the CSI labs and the business units. Customers are finding that their existing systems are more of a barrier than they expected. We can help our customers manage the object store technology and the data infrastructure, giving them more agility with their data.

Every part of the portfolio and research areas must be leveraged for an edge-to-cloud or, for industrial players, controls-to-cloud digital infrastructure that enables operations intelligence-style outcomes. If we can find a way to pick more of the semantics of the apps and put them into the infrastructure layers, the apps will become easier to develop. It will be faster and easier for customers to try new business models and be experimental about new ways to monetize their assets and data. By pulling together different kinds of data, different velocities and qualities and analyzing it in a cost-effective way, we create the outcomes they need. This is one way we can simplify the move to business model innovation for our customers.

Beverly
We take an innovative approach by bringing in industry experts – and not just technologists – to go out with the sales forces and enable conversations they couldn’t have had without them. With co-creation, we’re asking the customer what outcomes they’re looking for and work backwards from that.

Brijesh
The Social Innovation Business team engages with the customers to identify problems that potentially could give them a huge return on investment (ROI) and a competitive advantage. Usually we look for point solutions and we try to solve the problem as a first-of-a-kind solution using big data analytics. That’s a starting point for incubating the solutions.

Having the data infrastructure enable more DataOps, the way you get DevOps on the development side, is going to help address data engineering challenges and the big costs associated with wrangling the data. Moving from one-directional predictive maintenance and predictive analytics on sensor data, moving back into specific use cases where we can actuate a control – there’s big value there for the industrials. We want to be able to say we have this controls-to-cloud data infrastructure to enable your operations intelligence.

I’m going to add one more item and that’s repeatability. Customers are asking if we can make it repeatable from one solution to another. When moving from one use case to another, they don’t want to spend the same amount of money and effort starting over. They also don’t want to build another UI or dashboard. Essentially, it’s about repeatability.

Umesh
Absolutely. Repeatability of both the data flow and the analytics pipelines and then the dashboard, and then the controls, too. This is an area we’ve started to focus on in the research labs: How do we derive insights from data? What is the best action you can take and how do you turn these into actionable insights? In the case of a lot of these operational solutions, you have to be able to actuate so the question becomes: How do we marry AI machine-learning with controls? I think that’s where Hitachi has unique advantage because we have people with deep expertise on operational technology (OT) systems and we pair that with a lot of what we are doing on the machine-learning side.

Beverly
Absolutely. We have the subject matter experts and decision-support systems – the best and the brightest – and we are changing the systems. I’ve actually found two ways in which we are set apart from the industry no matter what internet of things (IoT) solution you are looking at. One is that we are humble. By this I mean that we don’t pretend to know all the answers. We tell the customer we want to co-create with you because we don’t know your issues and we want to take your knowledge and our knowledge and blend it together. The other differentiator is our humanness. We look at the human systems as well as the mechanical systems and we look at the processes and what the impact is on the system.

Umesh
The Insights Lab has become a center of excellence for analytics and machine learning and AI. It is important that we maintain our alignment with product and go-to-market strategies. We’ve talked about several areas: making the data pipeline work out of the box for solutions we’re building, driving down the application semantics into the infrastructure side and out into the edge. Other areas include supporting richer content types including video content so they’re easier to work with; and the coming together of AI and machine learning and controls and then scaling out to end-to-end optimization for complex multi-party systems. We’re looking at micro app architecture as well.

Brad
All of the pieces you mentioned aren’t siloed and that’s an important part as we work at this distributed architecture and data architecture that is evolving from hyperscale and hybrid clouds, and on-premises and edge-based storage. I think that also creates challenges for the data pipeline around, how to manage this distributed orchestration of data, the distributed ingestion of data, and the distributed analytics of the data on that edge-to-cloud or controls-to-cloud data infrastructure.

Brijesh
The business units want new use cases and solutions and help creating the new business models including big data analytics modeling support. The proof of concepts, or PoCs, that CSI is conducting are all customer-related. That means there was a need in the marketplace. Although they are first-of-a-kind projects, we have done the market research to know it has market potential. As soon as we start doing an incubation project, we bring in the business unit and get them involved whether it’s for the delivery of the solution or participating in the incubation itself. Also, it is important for research not to be isolated from the marketplace. We are taking the researchers to the customer and have them listen to what the customer needs.

Beverly
I’m looking for what the next best thing is that can be commercialized; solutions that are repeatable and will draw customers. We need market agreement, we need integrators, and we need everyone working as an ecosystem in order to be effective.

Brad
The more agile we can be; the better. If we make this a real partnership model working with research, we can work together around product strategy. Look at how the market is evolving, what are some of the technical differentiators today and where those trends are going, and then see how we can be on the leading edge of those trends by investing in those technologies and having the flexibility to adapt as we see new opportunities and solutions in the market start to really take off. We need to have a long-term vision for our products and where they’re going and make the right kind of technology investments to maintain a lead, but we need to be agile enough to pivot as we see an opportunity in the short term, and to adapt to the market as it changes.

Umesh
We have the technologies and the methods to extract the knowledge directly from the data and then automate. It’s not just about ingesting data and providing intelligence for human decision making. We want to take that and write it back into the control systems. We’re also exploring open innovation by reaching out to some of our research collaborators at universities and national labs to see if we can build on some of the work being done there and provide our practical experience, building out some of these solutions into the research programs.