posted

0 Comments

Here’s an excerpt from, ‘A Revised View of the IoT Ecosystem’, Vinton G. Cerf, IEEE Computing Edge, July 2018, (also appearing in IEEE Internet Computing, vol. 21, no.5, 2017) Emphasis mine.

 

  • For a long time, I had the idea the configuration of a suite (ensemble)** of Internet of Things (IoT) devices would be an infrequent process […]. However, my thinking has evolved. Sticking with the residential paradigm for a moment, although the ideas seem equally applicable to industrial settings, it’s becoming clearer that many devices will come and go with the residents, guests, workmen, emergency services personnel, and other who might have reason to enter the premises. […]. This leads me to believe that an IoT ensemble must actually be in a kind of continuous configuration mode, anticipating the arrival and departure of all manner of Internet-enabled devicesAmong the implications is the notion that the local IoT management systems needs to expect that new devices will need to be configured into the system and others to depart – it needs to sense their arrivals and departures and to react accordingly.

 

Vinton’s observation/conversion now matches how I’ve thought about the configuration, registration, de-registration and topology of sensors/actuators, IoT Gateways, and Datacenter Applications for some time. For the purpose of this article, consider the following topology, some number of sensors/actuators of low compute capacity, if any, connected by various local-area network protocols (e.g., CAN bus) to a system known in the industry today an IoT Gateway, which, in turn, is connected, by typical wide-area network protocols, to applications (management, analytics, business) running on server-class systems. An IoT Ensemble comprises all of the sensors, actuators, networks, gateways, and applications working together to produce a result valuable to organizations, society, and individuals.

 

It is clear that for some use cases, the relationships between the instance comprising of an IoT ensemble will remain fixed for long periods (years). For example, consider a legacy factory with a fixed set of controllers and machines where an IoT solution is collecting metrics on the behavior of the machines and controllers for the purpose of predicting problems requiring maintenance. In this case, the sensors, actuators, gateways, and applications need not change very often.

 

On the other hand, consider an indoor real-time location service consisting of small, battery powered tags, gateways, and a datacenter application. The gateways use triangulation of signal strength to determine the location of a tag. Tags can be associated to people and equipment and thus the application can tell a user immediately the location of any tagged piece of equipment or person. Clearly the relationship between a tag and a gateway changes reasonably rapidly as the person or equipment moves around the monitored space.

 

A good example use case comes from the healthcare industry. Studies have shown that there is quite a bit of productivity loss caused by personnel searching for equipment. An indoor, real-time, location service will certainly save much of the productivity loss. As an aside, the Gartner 2018 Hype-Cycle for IoT indicated that indoor, real-time location services is only one of two (the other being SmartLighting) that will reach plateau within the next two years.

 

Another use case requiring dynamic relationships between the components of an IoT ensemble is interesting. Utilities are looking at installing gateways on service vehicles (e.g., trash pickup trucks) that connect to smart energy meters (residential and commercial) to upload data from the meter and download updates. Clearly, again, the relationships among the components varies rapidly.

 

The bottom-line is that, to be applicable across many use-cases, the configuration, registration, de-registration, and connectivity among the components of IoT ensembles must be designed to allow rapid changes in the relationships. Implicit in this is that security consideration must also accommodate the dynamicity.

 

About Vinton G. Cerf

***Vinton G. Cerf, he’s one of the people who ACTUALLY invented the Internet. A very influential person for decades. He’s currently VP and Chief Internet Evangelist at Google, past president of ACM. He’s widely known as one of the ‘Fathers’ of the Internet. He is a Fellow of ACM and IEEE. https://en.wikipedia.org/wiki/Vint_Cerf