Infrastructure, Then Integration

The Internet of Things will roll out according to application first. After that, it’s all speculation.


In any business and in any business segment, you have to start somewhere and then figure out what the likely end points will be and how to get there. It’s no different with the Internet of Things, except that the integration of different markets or applications is rather fuzzy at the moment.

For example, it’s pretty easy to predict the possibilities for a car. It will likely move from connected infotainment to a self-driving device sometime in the next couple decades, with the time depending both on safety testing, volume and the ability of on-board diagnostics to predict failures before they occur and prevent a carefully orchestrated grid of cars from devolving into gridlock. The cars also will have to be secure from hacking, designed to prevent interference of different components—inside and outside of cars—and they will have to be highly energy efficient.

It’s also relatively straightforward to figure out possibilities for automation in industry, where uptime is critical and energy costs are operational expense line items. The industry defines the requirements and the needs, and the electronics are modified with enough sensors and know-how to maximize efficiency and prevent downtime.

Even the home market is somewhat predictable. Appliances and electronics can be part of an integrated network controlled by a smart device and secured from outside intrusion. They can minimize energy costs by adding more granularity into systems—when to charge a car or run a dishwasher, for example, or which parts of a home to heat or cool and which ones need to be the warmest or coolest.

What’s far less clear, however, is how these various segments will intersect, and what’s needed to both facilitate integration in some areas or how to protect them in other areas. The Internet of Things will begin life in separate markets, each with requirements, benefits and risks that are specific to those markets. So for medical electronics or aerospace, the requirements will be far different than for consumer electronics. Electronics have to last longer, have failovers, and they need to be more consistent and reliable for longer periods of time.

But how will these segments ultimately merge? And what are the ramifications when they finally do? That will take time to decide. There are proponents of a grand architecture that establish a set of rules, protocols and security requirements for the entire IoT. While that may sound logical enough on the surface, this kind of all-encompassing, empirical solution simply won’t work. Different markets will progress at different speeds, depending upon their needs, research and funding, and the competitiveness of those markets.

Moreover, not everything is quantifiable, meaning there is more resistance to adoption of new technology in some markets even where it may make sense. Consider the medical and legal markets, for example. Even after PCs were ubiquitous, most of their work was scribbled down by hand. Yet they now are avid users of technology. Medical will adapt to the IoT relatively quickly because of the benefits of non-intrusive and lower cost medical devices, but how they connect with other markets is likely to be far slower than consumer electronics.

As with any technology, markets dictate the speed of adoption. What remains to be seen is how multiple different markets play together in a giant ecosystem where the only common thread is technology. This is a new concept—talked about in the past but never implemented en masse—and it is likely to unfold in fits and starts, using plenty of caution in the beginning.