IoT development is often compared to technology precursors such as personal computers. While some of the comparisons make sense, there’s one reality that makes the IoT fundamentally different from technology trends preceding it: The “things” of the IoT will never coalesce into a single category, or even a handful of categories.
What’s the point of this distinction, and why should manufacturers of Industrial IoT products care? It’s because they need to make sure that they build their IIoT products to take into consideration this fragmentation — and the integration, compatibility, and complexity issues it entails, both now and for the future.
Computers and smartphones and tablets — the devices used to connect to the “Internet of Computers” — are designed with connectivity in mind. Not so with the products now being connected to the IIoT, which can range from lighting and turbines to boilers, thermostats, and farming equipment.
The disparate types of products connecting to the IIoT have vastly different ways to connect, require different networking protocols, have different purposes, and will generate different kinds of data. And these differences won’t somehow soften or disappear over time. A gas meter will always have different connectivity parameters than a tractor.
Still, for the IIoT to deliver meaningful value, disparate connected devices will need to coordinate their communications and — perhaps most significantly — the data they produce. At the end of the day, it’s this data where the real worth of the IIoT lies.
Here are some questions that manufacturers of IIoT products should ponder, ideally at the outset of the connected product design process:
- How will security be handled across the full spectrum of IoT connectivity, from the product through the cloud and to the software used to control the connected products?
- What kinds of networking protocols are needed to connect the products?
- Will protocol support be hard-coded into embedded firmware? If so, what happens if new protocols emerge in the future, or current protocols are no longer in use?
- What cloud infrastructure will the products connect to? Which services offered by the chosen cloud service provider will enable the most efficient and cost-effective operation of the connected products?
- Given that most IIoT products maintain persistent connections, how will those persistent connections be managed at scale so that cloud services expenses don’t skyrocket?
- Similarly, how will the data infrastructure be designed so that IIoT products’ potentially constant stream of data can be managed appropriately — to enable manufacturers to collect, organize, and make sense of the data?
- What will happen to a connected product’s operation when the cloud connection is lost? Specifically, how will schedules and other instructions be preserved when the Internet connection goes down?
- Will over-the-air (OTA) updates be offered? How will networking and security be handled for OTA communications?
- How will the connected devices be controlled? Through a web application? A mobile app?
- How will responsive performance be provided, so that when a user of the connected product initiates an action, the product responds immediately?
- How will security, performance, and reliability be scaled with increasing numbers or families of connected products?
- How will ongoing operation, maintenance, service, and support be handled for deployed IIoT products, which could easily remain in the field for years or even decades?
Unless an IIoT product manufacturer has deep technical expertise in all of these areas, it will be necessary to hire new personnel with the required skills, to outsource or acquire some or all of the skills, or to purchase a comprehensive IoT platform.
Whatever decision is right for a particular manufacturer, one thing is for sure: It’s not worth waiting with the hope that the IIoT will magically become less complex. The IoT will always be fragmented. Tapping the expertise of those who have already worked through the issues of fragmentation and complexity could ultimately save a significant amount of time, expense, and stress in the long run.
Originaly posted on Industrial IoT/Industrie 4.0 Viewpoints