Challenge: Keeping the Internet of Things grounded in reality

The IoT potentially will reach into everything, but enterprises need to zero in on where the most value will be realized.
Written by Joe McKendrick, Contributing Writer

In a recent post, ZDNet colleague Ken Hess wondered if the emerging Internet of Things may be the "Y2K" of our time. But he isn't concerned about IT systems collapsing in on themselves as a result of unaccounted code. Rather, as happened in the 1990s, he sees consultants, vendors, and self-proclaimed experts piling on to create a big carnival cashing in on business's fear, uncertainty, and doubt.

In some ways, the promises of IoT may be over the top. The idea that a flight delay will be communicated to someone's alarm clock as well as coffee maker, which will recalibrate to their owner's revised wake-up time is really weird stuff. Any IoT chatter that involves networks talking to a home refrigerator and automatically ordering more milk also may be more than the world needs.

IoT is one big, complex scenario, and really needs to be toned down to specific functions and applications where it really makes sense, and is straightforward to implement. There are some great feet-on-the-ground examples of where it is creating new business opportunities. In insurance, for example, auto insurers are installing telematics sensors into policyholders' cars (with their consent, so far) to track driving patterns — and offer discounts to good drivers.

GE, for its part, has taken a strong thought-leadership role in this space, observing that the so-called "industrial internet" will enable the widespread instrumentation of machines, ranging from aircraft engines to power plants to factory-floor production systems. These connected machines can let GE or another maker know when parts are wearing down, or provide paths to efficiency and energy savings.

Good stuff — and certainly the next phase of human progress. And it's about big data — there will be a lot of it — and having the processing power to handle the analytics (calling cloud). While it's proven we have the technology to do a lot of the IoT things we're hearing about, perhaps we need to take a step back and weigh its implications on other fronts.

Andy Oram, writing in O'Reilly Radar, provides some thought-provoking questions about the IoT, particularly around the way data is collected and handled. These issues were raised at the recent IoT Festival held at MIT — attended by systems developers, security experts, data scientists, and artists from across a spectrum of industries. Many of the hurdles are social, not technical, Oram says.

Questions that need to be asked by enterprises as they move into an IoT world include examining what data needs to be collected and managed. Oram observes that in today's big data world, we are already awash in large volumes of data, and enterprises are struggling to identify what nuggets of it are of value. Plus, a lot of the world's data is still locked away and  inaccessible.

"How much can we trust the IoT? How much can we take humans out of the loop?" For example, Oram illustrates, we need to ask if we can trust IoT-based controls to autonomously drive our cars, especially if we are moving at 65 miles per hour down a busy highway.

Plus, there were questions about the effects of all this data collection, and the injection of intelligence into devices, on privacy and personal autonomy. This already-burning question in the big data analytics realm regarding personal communications will continue to be an issue. There are trade-offs involved. Another question that came out of the IoT Festival is "How much privacy and personal autonomy are we willing to risk to reap the IoT’s potential for improving life, saving resources, and lowering costs?" Oram relates. The trade-offs are between privacy and smarter networks.

And, as is the case with everything in the IT world, there are issues with standardization and interoperability. As IoT takes hold, there will be issues in  persuading manufacturers to build standard communication protocols into everyday objects, says Oram. Industry standards are always problematic, even when eventually smoothed out as time goes by. But that often takes years.

(Thumbnail photo: Joe McKendrick.)

Editorial standards