It is highly likely the Internet of Things (IoT) will transform manufacturing – and that open source communities will play an important role as
part of the transformation.
In my most recent blog post, I wrote about how open source software has a role to play in manufacturing enterprises, especially in areas like machine learning and data analysis. Open source communities are at the leading edge of problem-solving using big data within this field.
Now I want to address the role of open source in an even bigger data issue—the Internet of Things. As vast as the data being generated by today’s factories may seem, it’s nothing compared to what is coming with the IoT.
How big will it be? According to Gartner, there are about 5 billion “things” connected to the Internet today, and that number will grow to 25 billion by 2020. The economic impact will be huge, too. McKinsey estimates that the IoT will bring $4 trillion to $11 trillion of positive economic impact each year by 2025.
What Does This Have to do With Manufacturing?
In manufacturing, the Industrial IoT or IIoT will bring monumental changes in how executives see and control their enterprises. (Editor: Point of clarification: Manufacturing Transformation sees the IoT as the “macro” term describing all interconnected devices; the IIoT refers to just the “things” that are specifically part of the manufacturing process.)
The recent advances in traceability are a good benchmark to consider. Enterprises that have global genealogy and traceability can pinpoint the source of problems down to a specific line and shift, thus limiting harm and the cost of containment. But current technology has only scratched the surface. Within a few years, everything that happens in manufacturing will be known—theoretically at least. Every wrench, widget, device, machine, label, truck, badge, scanner, and loose bolt will be linked in what will amount to a “neural network” of the complete production process. Manufacturers won’t have to wait until there’s a problem to contain it. They’ll see it before it happens, through the billion eyes of the IIoT.
That is just one example. The same kind of data-based applications will be enabled literally everywhere in the manufacturing process, across all disciplines, organizations, and facilities.
If all the data can be captured and managed. And if the tools exist to sort through this vast gold mine of data in real time, identify the nuggets of value, and deliver them to the right people.
Those are big ifs, and to solve them we will need open source software along with the communities that support them.
Challenge 1: Getting Everything to Interoperate
The first challenge is fundamental: interoperability. Ironically, the problem right now is that we have too many standards. There are more than 400 standards for “thing-to-thing” communication.
If one manufacturer’s building sensors use a standard that doesn’t communicate with another manufacturer’s process instruments, then this will obviously lead to issues. Imagine a supply chain scanner that only worked in some warehouses. This incompatibility would eliminate synergies across the enterprise, leaving only siloed systems and islands of information. This scenario would block the whole idea of the IIoT, which is to communicate with everything, everywhere.
Open source may hold the answer. Companies tend to gravitate toward open standards, rather than being tied to proprietary ones. The huge diversity of devices and items joining the Internet of Things will help drive the market toward open systems, in much the same way that web sites gravitated to HTML. You can build your web site in something else, but why would you?
An example of an open source standard within the world of web site integration is Node.js, which allows the creation of web servers and networking tools, using JavaScript and a collection of “modules” that handle various core functionalities (source). Node.js is based on an event-based architecture. It is very well suited for high-end user interfaces and a positive user experience.
Challenge 2: Storing Data
If everything is intelligent and connected, the amount of data that potentially could be captured in a single production facility is staggering. Even today, it isn’t possible to capture every possible data point at every moment. Consequently, big data experts have developed methods to sample and screen big data, saving only what is needed. That still amounts to petabytes of data, and most of the impact from the IoT is yet to come.
Here too, open source software, along with the cloud, points the way. A good example is Apache Hadoop, an open source framework for distributed storage and processing of big data, using computer clusters built on commodity hardware.
Looking forward, we can expect that this kind of solution, operating in the cloud, will provide all the storage needed at a practical cost no matter how big the IoT becomes.
Challenge 3: Using the Data
This is the real challenge. McKinsey estimates that only 1% of the big data collected by companies is put to actionable use. One reason is that companies have been focused on identifying anomalies, and haven’t begun to tap the potential for optimizing and advancing processes using big data.
That will change as industry leaders see the value and begin to leverage the gold mine of data they own. Not surprisingly, the open source world will be key here as well.As I wrote in my last blog, open source communities are at the forefront of application development in fields like machine learning, pattern recognition, and analytics that will be crucial for leveraging the big data of the IoT.
Naturally, we cannot be sure what changes the Internet of Things will bring. There are bound to be unexpected capabilities and benefits. One thing does seem certain – future advances and breakthroughs will likely be done by groups vs. individuals, as new best practices become apparent. Given the typical profile of open source communities, my bet is that this group of enthusiastic and passionate experts will be on the forefront of the evolution, and we should welcome their support with open arms.
If you liked this article, here are others you might also find interesting: