Thinking Big About The Industrial Internet of Things

The Industrial Internet of Things (IoT) is a network of physical objects imbued with information and communications technologies. It brings together many of the key technologies that I have argued will make or break every information intensive company.

McKinsey estimates that IoT will have an potential economic impact of up to $6.2 trillion by 2025 and the potential to drive productivity across $36 trillion in operating costs across multiple industries, including manufacturing, health care, and mining.

Jeff Immelt, chief executive of General Electric, shares McKinsey's optimism. Immelt called IoT “beautiful, desirable, and investable.” GE estimates that a 1% improvement in productivity across its global manufacturing base translates to $500 million in annual savings. Worldwide, GE thinks a 1% improvement in industrial productivity could add $10 trillion to $15 trillion to worldwide GDP over the next 15 years.

Brian Carpizo leads the manufacturing and supply chain team at Uptake, a Chicago-based company that partners with large industrial companies to leverage IoT opportunities. Uptake was named by Forbes as the hottest startup in 2015. It was described by Fortune as “saving industrial giants from disruption.” (Disclosure: I have been an advisor to Uptake.)

I recently explored with Brian how manufacturing giants can think big, start small and learn fast to address the disruptive innovation opportunities and challenges afforded by the Industrial Internet of Things. His observations are also relevant, I believe, to companies in other industrial sectors pondering their IoT strategies.

Chunka Mui: Let’s set some context. How are you thinking about the Internet of Things (IoT)?

Brian Carpizo: I’m focused on industrial insight in manufacturing enabled by IoT. Innovations in hardware, connectivity, big data, analytics, and machine learning are converging to create massive value-creation opportunities.

Here are the key elements: We can generate and capture valuable machine, operational, and environmental data. We can store, retrieve and process massive data sets from disparate sources—including the silos that separate data-center-based (IT) systems and plant-oriented operational technology (OT) systems. And, we can turn that data into operational insight. Rather than one-off exercises in the lab, we’re building persistent tools that improve daily operations.

Thinking Big

Mui: You said that you are focused on manufacturing. Help us understand the big picture. How might IoT transform manufacturing?

Carpizo: IoT enables three transformative opportunities in manufacturing: machine and factory health, digital thread, and smart products. Let me briefly explain each of these:

Machine and factory health: Advanced automation, complexity, and lean supply chains increase the probability that things can go wrong in manufacturing settings—and the cost of those disruptions are massive. We use data mining techniques when processing signals streaming from machine sensors to detect machine degradation and predict impending failures.

Digital thread: We are analyzing converged IT/OT data sets to identify previously hidden correlative factors in the entire product lifecycle. We are tearing down the silos between design engineering, manufacturing engineering, manufacturing operations, and after-market support. We treat it as one process to be tuned and optimized. Instead of just look-behind tools, we predict how changes in inputs and processes can affect output.

Smart products. We are placing intelligence into manufactured products to expand core value propositions. Smart, connected products provide OEMs operating data and enable two-way product support. For example, imagine the value of correlating manufacturing inputs and processes to spikes in warranty claims or product defects automatically detected when the product “phones home”. Connected products can also inform the supply chain about things like predicted demand for aftermarket parts.

Mui: How do you see these opportunities affecting the competitive dynamics between manufacturers?

Carpizo: We’re still in the early stages of IoT in manufacturing so there are definitely opportunities to obtain first-mover advantages.

Internally, there are cost, throughput and quality improvement opportunities. Customers might not readily see these innovations, but they will strengthen those that excel at the expense of those that don’t.

Externally, there are opportunities to create additional value beyond the hardware. For example, the Nest thermostat demonstrates that if you build intelligent control into a relatively commoditized product, you can extract higher margins, expand the market, and generate enterprise value.

Just like other waves of technology-enabled disruption, this will create a new category of leaders, followers and laggards—and the market will reward and punish accordingly.

Starting Small

Mui: You have been working in this area for several years. What are the biggest challenges for manufacturers as they begin to focus on this topic?

Carpizo: Data integration and system interoperability is the first challenge. There are a lot of systems and machines producing data but very few ways to exchange and store information in the standardized way required for adoption of IoT solutions at scale. The state of today’s IT/OT integration is very one-off, labor intensive, complex, and inflexible. And, we still have yet to develop a universal “machine data language” that promotes plug-and-play solutions.

A similar problem in the 90s in both the IT and OT worlds triggered the evolution of applications into integrated application suites like ERP and integrated control systems. In contrast, the converged IT/OT world does not lend itself to one-vendor systems of record or some kind of mega-ERP. The problem is just too complex.

The plumbing effort to put IT/OT together must become easier if we are to see adoption at scale —and that’s something that will be driven by new data services solutions combined with an industry effort to coalesce around data standards like MT Connect, OPC-UA, and the recently announced Open Connectivity Foundation. I predict we’ll see major advancements in this area in the next few years.

Mui: What is the best way to get started?

Carpizo: There are a few approaches. One is the pilot approach; characterized by small scope but a relatively deep set of use cases. This approach might involve a lower cost, throwaway architecture that while not scalable, secure or fault tolerant, helps inform the future strategy with real experience.

The other is a bolder approach—a big bang type initiative that re-defines the company strategy. Here, Caterpillar, an Uptake partner, is a great example. The bolder approach is appropriate when the benefits are well understood, sizable, and attainable. This requires that the CEO and senior management are on board and are driving the strategy since it requires a high degree of cross-functional coordination.

Mui: Manufacturing runs on a much slower clock speed than other industries being disrupted by digital technology, like retail, entertainment, and consumer products. How does that impact efforts?

Carpizo: The age of legacy technology is critical. Old technologies in fast-paced industries typically die more quickly. In manufacturing, it isn’t uncommon to have 15-20 year old systems in either IT or OT. These “zombie” systems exist because they work OK and there is no burning platform to make companies upgrade. So the lack of ability for some of these older systems be a part of a “connected enterprise” can impact efforts.

The benefits and acceptance of IoT solutions will force an upgrade cycle in the coming years—and the ability and agility to respond will be one key factor that separates the winners from the losers.

Mui: What else might impede progress?

Carpizo: Leaders must recognize that substantial IoT initiatives can be incredibly disruptive. It is likely the organization will have to be realigned to support cross-functional decision making. Be prepared for the possibility that IoT will enable entirely new lines of business, fundamentally change distribution channels, and forge new strategic alliances. Success will depend on clear senior executive mandate and shelter from organizational resistance—while ensuring that critical stakeholders get on the same page in understanding and supporting the effort.

Learning Fast

Mui: A demo is worth a thousand pages of a business plan. What are the most dramatic demonstrations of the potential of the IoT and predictive analytics in manufacturing?

Carpizo: In manufacturing, seemingly identical processes can produce big variations in output. For example, injectable drug and biologics pharmaceutical manufacturing appears to be a highly-controlled closed system with very tight quality requirements. On closer examination, however, there are differences in hundreds of variables like the rate and order of addition of the ingredients to the solvent, the rate of temperature change, the RPMs of the impeller in the mixing vessel, the position of the blades in the mixing vessel, the formulation makeup of the solvent, etc. Running machine learning algorithms over data from multiple batches over all the data will identify which of those variables are impacting the output characteristics. Some of the data will come from the formulation management system, some will come from the manufacturing execution system, and some will come from the plant machinery.

This kind of result can be a real eye opener for product managers, process engineers, QA and plant managers. They are so close to the process and often assume they know everything there is to know about how to best make a product, but there are limitations to how much data and how many simultaneous variables a human can consider. This is a case where machine learning augments human expert knowledge.

Mui: What lessons can be drawn from early efforts?

Carpizo: If you talk to early adopters and follow the latest developments, you will find ample evidence from studies, pilot projects and high-value deployments on large pieces of machinery.

The challenge now is to move forward from these efforts and scale deployment. We are starting to tie things together to promote cross-functional insight and close the loop between design, manufacturing, and maintenance.

Mui: What are the tough questions manufacturing industry Boards of Directors and senior managements should be asking about the IoT right now?

Carpizo: Although there is a lot of optimism about the potential impact of IoT, most companies are sitting on the sidelines and waiting for the picture to clear up. The tough questions include: What is the opportunity cost for waiting? What is the impact of ceding to a known or unknown competitor an opportunity to stake out this ground? How can we influence the changing industry dynamics instead of reacting to what others are doing?

--

Chunka Mui is a business advisor and author of three books on strategy and innovation including, most recently, The New Killer Apps: How Large Companies Can Out-Innovate Start-Ups.  This article was originally published at Forbes.