Parisi.IO

Loading...Reticulating Splines


Home

Navigating IoT

Navigating the Internet of Things.

The Internet of Things is a vague description that encompasses all devices that could possibly be networked. If a device has a networking card or a radio, it is a thing on the internet. In some cases, it may be confusing to know what IoT means because we really can’t exclude any networked device from this description.

In search of the etymology of the initialism I found myself rummaging through many opinions on the origin [1]. All correct, I’m sure, but what it leads me to believe is that we are free to define IoT in the realm of our organization. IoT may or may not be important to one’s job, but it’s important to recognize that all devices, phones, cars, remote controls, etc create data and send it cross the internet.

Projects like the one I work on, Apache MiNiFi, aim to corral data generated by these devices; however, the question remains “what do I do with it?” In many cases these devices are single purpose collectors and provide little utility outside of their intended use case. In other cases, the IoT devices may have significant compute capacity allowing these devices to be remote haven of code execution. What I find perplexing is: does the IoT revolution mean we are supposed to buy more devices or use the ones we already have?

IoT is a very simple term by nature, but the advertising implies that we somehow need to purchase a service from some of the big players to use these devices; however, in many cases protocols already exist. Philips Hue devices already have a protocol and communicate with your hub and thus some centralized service. Only in rare cases will custom services or software be needed. If you are building a new device, it may be advantageous to use software to corral your data if it means reduced development time and improved provenance of data. What doesn’t make sense is investing resources in massive infrastructure and complicated software.

Be mindful that getting data from your devices may not dramatically improve your overall system experience. Being at the front of the IoT revolution may mean that you can build more business intelligence, but it may also mean you spend significant time and money to get very small amounts of meaningless data. Like with all fads, just because there is data to be obtained, doesn’t mean the return on investment is there.

The beauty of IoT being as general as it is means that the paradigms may apply to infrastructure that is within a controlled environment. Perhaps it means you can get the status of manufacturing hardware sooner, and apply machine learning to estimate hardware failure before it occurs. I visited a ‘big data’ conference in the Midwest and listened intently to a company that did just that in the IoT track. Their presentation was one of caution as they said that the companies never shut down the equipment based on a warning, even after the software repeatedly identified premature failure. Listening and repairing would have saved time and money, but the companies simply stayed with their ways. The IoT revolution didn’t revolutionizing manufacturing for this company because IoT, while a penetrating paradigm, wasn’t a panacea to everyone.

Much like the ‘big data’ movement, technology is only meaningful when it is something everyone sees as a commodity. Do you need IoT or ‘big data’ to perform your work? For me that means trusting the results and information enough to accept downtime or accept cost; however, for everyone this clearly won’t be the case. When you navigate IoT, it’s important to consider that technologists are applying a nomenclature referring to the connectedness of devices that typically weren’t connected previously. It doesn’t necessarily mean that you will have lower costs or more information, and even if it does, it doesn’t mean that your organization will trust that information.

Will IoT spell greater medical breakthroughs or manufacturing improvements?  I don’t think by itself; however, we don’t always know what to expect from our infrastructure. At this same conference, I heard a talk by a veterinarian on medically linked devices for pets. These helped predict health markers that indicated the health of an animal that can’t speak or provide details. As we imagine what is possible, more devices will be created that leverage the IoT paradigm.

Security will always be a concern of mine, so I don’t anticipate this ever going away. Since the internet of things brings things together that weren’t linked before, we have risk that everyday devices fail due to vulnerabilities or provide PII without our permission. IoT can mean great utility, but it can also mean major security vulnerabilities. I would caution that advancement of technology should never take a back seat to security; however, for this reason I do opt to not include smart locks of thermostats to my home. I have smart lights, which can provide a threat vector into my home network; however, I don’t want to have someone else potentially control my thermostat or someone enter my house with a simple Bluetooth hack. When I look at the IoT realm I see many possible devices and the security threats limit what I am willing to accept; however, it is quite possible that we can mitigate these concerns when leveraging infrastructure in an organization.

In conclusion, the Internet of Things, is an interconnected network of a cacophony of devices, some old and some new. What’s important is that these devices, often single or few purpose, provide meaningful data to some, but a flutter of noise to others. You must evaluate what is important to your organization. Having an IoT infrastructure and collecting massive amounts of data may not be ultimately useful for your company. The security risks should also be imagined and you must determine if the return on investment is worth any risk. There are many applications from manufacturing to medical, and anywhere in between. Navigating what will be useful and worth the cost must come from honest discussion about whether your organization is willing to make change from the data collected. Otherwise, IoT is just noise.

[1] https://www.redbite.com/the-origin-of-the-internet-of-things/