Design, deliver, and run enterprise blockchain workloads quickly and easily.
Why effective IoT adoption is a team sport
Implementing an Internet of Things (IoT) strategy is not a one-size-fits-all endeavor—nor can a solution be bought off the shelf. What’s more, those new to the computational hive and analytical edge attributes of IoT are discovering that it takes a team approach. This Q&A, also available as a podcast, examines how the complexity and novel architectural demands of IoT require a rethinking of the edge of nearly any enterprise.
Dana Gardner: To explain how many disparate threads of an effective IoT fabric come together, we're joined by Tushar Halgali, senior manager in the technology strategy and architecture practice at Deloitte Consulting.
Tushar Halgali: Dana, thanks for having me.
Gardner: We're also here with Jeff Carlat, senior director of technology solutions, Strategic Alliances, at Hewlett Packard Enterprise. Welcome, Jeff.
Jeff Carlat: Thanks. It’s nice to be here, Dana.
Gardner: What are some of the top trends making organizations recognize the importance of IoT? What’s making them examine the architectural implications?
Carlat: We're at the cusp of a very large movement of digitizing entire value chains. Organizations have more and more things connected to the Internet. Look at your Nest thermostats and the sensors that are on everything. The connectivity of that back to the data center to do analytics in real time is critical for businesses to reach the next level of efficiencies—and to maintain their competitiveness in the market.
Gardner: Tushar, this is a different type of network requirement set. We’re dealing with varied data types, speeds, and volumes in places that we haven't seen before. What are the obstacles that organizations face as they look at their current infrastructure and the need to adapt?
Halgali: One of the really interesting things we've seen is that traditionally organizations have been solving technology-related problems as all information technology-related problems. There was this whole concept of machine to machine (M2M) a while back. It connected machines to the Internet, but it was a very small segment.
Now, we're trying to get machines to connect to the Internet and have them communicate with each other. There are a lot of complexities involved. It's not just the IT pieces, but having the operational technology (OT) connect to the IT world, too. It creates a very complex ecosystem of components.
Gardner: Let’s parse out the differences between OT in the IT. How do you describe those? Why should we understand and appreciate how different they are?
Carlat: When we think of OT, you think of long-standing companies out there, Bosch, National Instruments (NI), and many other companies that have been instrumenting sensors for operations, shop floors, oil and gas, and with every pump being sensed. The problem is that humans would have to interact a lot around those sensors, to remediate or to understand when something like a bearing on a pump has failed. [Learn more on OT and IoT.]
What's key here is that IT, those core data center technologies that HPE is leading the market in, has the ability of run analytics and to provide intelligence and insights from all of that sensor data. When you can connect the OT devices with the IT—whether in the data center or delivering that IT to the edge, which we call the Intelligent Edge—you can actually do your insights, create your feedback, and provide corrective actions even before things fail, rather than waiting.
Gardner: That failed ball bearing on some device isn't just alerting the shop floor of a failure; it's additionally automating a process where the inventory is checked. If it’s not there, the supply chain is checked, the order is put in place, it’s delivered and ready to install before any kind of breakdown—or did I oversimplify that?
End of downtime
Carlat: That’s a fair representation. We're working closely with a company called Flowserve. We’re building the telemetry within the pumps so that when a cavitation happens or a bearing is starting to wear out, it will predict the mean time for failure and alert them immediately. It's all truly connected. It will tell you when it’s going to fail. It provides the access to fix it ahead of time, or as part of a scheduled maintenance plan, rather than during downtime, because downtime in an oil production facility or any business can cost millions of dollars.
From the Intelligent IoT Edge
Gardner: Tushar, are there any other examples you can think of to illustrate the power and importance of OT and IT together?
Halgali: If our readers ever get a chance to check out one of the keynote speakers [at HPE Discover London 2016] on the Intelligent Edge, there's a good presentation by PTC ThingWorx software, which is an IoT platform and the HPE Edgeline servers in a manufacturing facility. You have conveyor belts that need certain improvements; they're constantly producing things, and they're part of the production facility. It’s all tied to the revenue of the organization, and the minute it shuts down, there are problems.
Maintenance needs to be done on those machines, but you don’t want to do it too soon because you're just spending money unnecessarily and it’s not efficient. You don’t want it too late, because then there's downtime. So, you want to find the equilibrium between the two.
IoT determines the right window for when that maintenance needs to be done. If there's a valve problem, and something goes down quickly, sensors track the data and we analyze the information. The minute that data goes off a certain baseline, it will tell you about this problem—and then it will say that there’s the potential in the future for a major problem.
It will actually generate a work order, which then feeds from the OT systems into the IT systems, and it’s all automatic. Then, when mechanics come in to try to solve these problems, they can use augmented reality or virtual reality to look at the machine and then fix the problem.
It’s actually a closed-loop ecosystem that would not have happened in the M2M base. It’s the next layer of maturity or advancement that IoT brings up.
Gardner: We can measure, we can analyze, and we can communicate. That gives us a lot of power. We can move toward minimum viable operations, where we're not putting parts in place when they're not needed, but we’re not going down either.
It reminds me of what happened on the financial side of businesses a decade or two ago, where you wanted to have spend management. You couldn't do it until you knew where all your money was, where all the bills had to be paid, but then doing so, you could really manage things precisely. Those were back-office apps, digital ledgers.
So, it’s a maturity coming to devices—analog, digital, what have you, and it’s pretty exciting. What's the impact here financially, Jeff?
Carlat: Well, huge. Right now, IDC predicts IoT to represent about a $1.3 trillion opportunity by 2020. It's a huge opportunity, not only for incremental revenue for businesses, but increased efficiencies, reducing cost, reducing downtime, reducing risk—so, a tremendous benefit. Companies need to strongly consider a movement for digitizing the value chains to remain competitive in the future.
Bigger and better data at the edge
Gardner: OK. We understand why it's important, and we have a pretty good idea of what you need to do. Now, how do you get there? Is this big data at the edge? I heard a comment just the other day that there's no bigger data than edge data and IoT data. We're going to have to manage scales here that we haven’t seen before.
Carlat: It’s an excellent point. Jet engines that are being used today are generating 5 TB of data every time they land or take off. Imagine that for every plane, every engine that’s flying in the sky, every day, every year. The amount of data is huge. This brings me to the unique way that HPE is approaching this, and we truly believe we are leaders in the data center now and are leaders within IT.
We're taking that design, that implementation, that knowledge, and we're designing infrastructure, data center quality infrastructure, that’s put on the edge, ruggedized compute or analytics, and providing the ability to do that analysis, the machine learning, and doing it all locally, rather than sending all that data to the cloud for analytics. Imagine how expensive that would be.
That's one approach we're taking on within HPE. But it’s not just about HPE tackling this. Customers are asking where to start: "This is overwhelming, this is complex. How do we do this?" We're coming together to do advisory services, talking our customers through this, hand-holding, building a journey for them to do that digitization according to their plans and without disrupting their current environment.
Gardner: Tushar, when you have a small data center at the edge, you're going to eke out some obvious efficiencies, but this portends unforeseen circumstances that could be very positive. What can you do when you have this level of analytics, and you take it to a macro level? You can begin to analyze things on an industry-level, and then have the opportunity to drill down and find new patterns, new associations, perhaps even new ways to design processes, factory floors, retail environments. What are we talking about in terms of the potential for the analytics when we capture and manage this larger amount of data?
Halgali: We've noted there are a lot of IoT use cases, and the value that generates so far has been around cost optimization, efficiencies, risk management, and those kinds of things. But by doing things on the edge, not only can you do all of those, you can start getting into the higher-value areas, such as revenue growth and innovation.
A classic example is remote monitoring. Think of yourself as a healthcare provider who would not be able to get into the business of managing people's health if they're all located remotely. If we have certain devices in homes through sensors and everything, you can start tracking their behaviors and their patterns. When they're taking medicine and those kinds of things, and have all the information created through profiles of those people. You have now distributed the power of taking care of all the constituents in your base, without having to have them physically be in a hospital.
Gardner: Those feedback loops are not just one way where you can analyze, but you can start to apply the results, the benefits of the analysis, right back into the edge.
Carlat: Health and life sciences are great examples of using IoT as a way of more efficiently managing the hospital beds. It costs a lot of money to have people sit in a hospital when they don't need to be there. To be able to provide patient access remotely, to be able to monitor them, to be able to intervene on an as-needed basis, drives much greater efficiencies.
We’ve talked a little bit about industrial IoT, we’ve talked a little bit about health and life sciences, but this extends into retail and smart stores, too. We're doing a lot with Home Depot to deliver the store of the future, bridging the digital with the brick-and-mortar across 2,200 stores in North America.
It also has to do with the experience around campus and branch networks. At Levi’s Stadium in Santa Clara, California, HPE built that out with indoor Global Positioning System (GPS) and built out a mobile app that allows indoor wayfinding. It allows the patrons visiting the game to have a totally new, immersive experience.
They found uploads and downloads of photos, and they found hotspots by mapping out in the stadium. The hotspots had a great unobstructed view of the field, so there were a lot of people there taking pictures. They installed a food stand nearby, and they have increased revenues because of strategic placement based on this IoT data. Levi’s Stadium recognized $1 million in additional revenue in the first season and 10 times the growth in the number of contacts that they have in their repository now.
Gardner: So, it's safe to say that edge computing and intelligence is a gift that will keep giving, at levels organizations haven’t even conceived of yet.
Carlat: I believe it’s a necessity to stay competitive in the world of tomorrow.
Gardner: If your competitor does this and you don't, that’s going to be a big question mark for your customers to mull over.
While we are still on the subject of the edge technical capabilities, by being able to analyze and not just pass along data, it seems to me it's also a big help when it comes to compliance and security, which are big concerns.
Not only does security get mitigated by hardening or putting up a wall, probably the safest bet is to be able to analyze when something is breached or something is going wrong, and then to control or contain that. Tell me why the HPE Edgeline approach of analyzing data fast and on the edge can also be a big boost to security risk containment and other compliance issues.
Carlat: We do a lot around asset tracking. Typically, you need to send someone out there to remediate. By using Edgeline, using our sensor data, and using asset tagging, you can ensure that the right person can be identified as the service person physically at the pump to replace it, rather than just saying that they did it, writing on paper, and actually being off doing something else. You have security, you have the appropriate compliance levels with the right people fixing the right things in the right manner, and it's all traceable and trackable.
Halgali: When you begin using edge devices, as well as geolocation services, you have this ability to do fine-grained policy management and access control for not just the people, but also devices. The surface area for IoT is so huge, there are many ad hoc points into the network. By having a security layer, you can control that, and edge devices certainly help with that.
A classic example would be if you have a camera in a certain place. The camera is taking constant feeds of things that are going on that are wrong or right; it’s constantly recording the data. But the algorithms that have been fed into the edge device allow it to capture things that are normal, so it can not only alert authorities at the right time, but also store feed only for that. Why store days and days' worth of images, when you can pick only the ones that truly matter?
As Jeff said, it allows workplace restrictions and compliance, but also in an open area, it allows you to track events that are specific.
In other cases—let’s say the mining industry or the oil and gas industry, where you have workers that are going to be in very remote locations and it’s very hard to track each one of them—when you have the ability to track the assets over time, if things go wrong, then it’s easier to intervene and help out.
Carlat: Here is a great personal example. I went to my auto dealership and I pulled into the garage. Immediately, I was greeted at my door by name, “Hello Mr. Carlat. Are you in for your service?"
I thought, “How do you know I came in? Are you tracking me? How are you doing that?” It turns out, they have radio-frequency identification (RFID) tags. When you come in for service, they apply these tags. As soon as you pull in, they provide a personalized experience.
Also, it yields a net benefit of location tracking. They know exactly where my car is at all stages. If I moved to a rental car that they have there, my profile is automatically transferred over there. It starts their cycle time metrics, too, the traceability of how they're doing on remediating whatever my service level may be. It's a whole new experience. I'm now a lifetime-loyal customer of this auto dealer because of that personalization; it’s all coming from implementation of IoT.
Gardner: The implications are vast; whether it’s user experience, operational efficiency, risk reduction, or insights and analysis at different levels of an industry and even a company.
It's very impressive stuff, when you can measure everything and you can gather as much data as you want, and then you can triage and analyze that data and put the metadata out to the cloud; so much is possible.
We've established why this is of interest. Now, let’s think a little bit about how you get there for organizations that are thinking more about re-architecting their edge in order to avail themselves of some of these benefits. What is it about the HPE and Deloitte alliance that allows for a pathway to get on board and start doing things in a proper order to make this happen in the best fashion?
Transformation journey, one step at a time
Halgali: Dana, anytime you do an IoT initiative, the key thing to realize is that it should be part of a major digital transformation initiative. Like any other transformation story, there are the people, process, and the technology components of it. Jeff and I can talk about these three at a very high level when you begin talking about the process and the business model.
Deloitte has a huge practice in the strategy and the process space. What we're looking at is digital industrial value-chain transformation. Let’s look at something like a smart factory.
What’s the value chain for an organization that's making heavy machinery, end to end, all the way from research and development and planning, to procurement and development and shipment, and after-sale repairs, the entire value chain? What does that look like in the new IoT era? Then, decompose that into processes and use cases, and then identify which are the most high-value use cases, quantifying them, because that’s important.
Identifying the use cases that will deliver immediate tangible value in the near term provides the map of where to begin the IoT journey. If you can’t quantify concrete ROI, then what’s the point of investing? That addresses the reason of what IoT can do for the organization and why to leverage this capability. And then, it's about helping our clients build out the business cases, so that they can justify the investments needed from the shareholders and the board—and can start implementing.
At a very high level, what’s the transformation story? What's the impact on the business model for the organization? Once you have those strategy questions answered, then you get into the tactical aspects, which is how we execute on it.
From an execution standpoint, let’s look at enablement via technology. Once you have identified which use cases to implement, you can utilize the pre-integrated, pre-configured IoT offerings that Deloitte and HPE have co-developed. These offerings address use cases such as asset monitoring and maintenance (in oil and gas, manufacturing, and smart cities), and intelligent spaces (in public venues such as malls, retail stores, and stadiums), and digital workplaces (in office buildings). One must also factor in organization, change and communication management, as addressing cultural shifts is one of the most challenging aspects of an IoT-enabled digital transformation. Such a holistic approach helps our clients to think big, start small, and scale fast.
Gardner: Tushar just outlined a very nice on-ramp process. What about some places to go for information or calls for action? Where should people get started as they learn how to implement on the process that Tushar just described?
Carlat: We're working as one with Deloitte to deliver these transformations. Customers with interest can come to either Deloitte or HPE. We at HPE have a strong group of technology services consultants who can step in and help in partnership with Deloitte as well.
So, come to either company. Any of our partner representatives can get all of this and our websites are loaded with information. We're here to help. We're here to hold the hand and lead our customers to digitize and achieve these promised efficiencies that can be garnered from digital value chains.
Gardner: So the buzzwords to track or to follow to get into your organizations are Internet of Things, Operational Technology. Did I leave anything out?
Halgali: Digital Analytics, Edge Computing.
Gardner: I'm afraid we will have to leave it there. We've been exploring how IoT adoption means more than just scaling up networks. The complexity and novel architectural demands of IoT require a rethinking of the edge of nearly any enterprise.
And we’ve learned how to rationalize the many disparate parts to produce an effective IoT fabric through a team approach. So, thanks to our guests, Tushar Halgali, senior manager in the technology strategy and architecture practice at Deloitte Consulting, and Jeff Carlat, senior director of technology solutions at HPE strategic alliances.
This article/content was written by the individual writer identified and does not necessarily reflect the view of Hewlett Packard Enterprise Company.