In a Fog About the Cloud?
By Alan Dash
Alan Dash is senior advisor with Impact Advisors of Naperville, IL.
The use of a cloud to symbolize some magical spot where all the answers to the world’s questions are housed and an infinite amount of storage exists has been around since the 1970s. I reflect on my own career, while programming for the US Air Force in the early 1980s, drawing clouds in my diagrams to show that somewhere, out there, beneath the pale moonlight, someone’s thinking of me, and filling the void symbolized by my cloud with meaningful data.
Not exactly how Linda Ronstadt and James Ingram sang it, but that was my visual. Back then we called it what it was – centralized computing; output devices received data from centrally-located applications.
Then came PCs, placing those applications out onto the edge of the computing environment and away from the monster in the data center that threw off a dragon’s amount of heat and occasionally an equal amount of fire and brimstone. We called that de-centralized computing; everyone was free to process at their desk.
PCs became smaller, applications bigger. Soon we needed gigabits of storage to hold the very applications that were to be fed with an obese amount of data. Ultimately PCs couldn’t handle the power and space needed, so centralized computing came back, only this time we called it “The Cloud” and it was good – good because we learned new acronyms like SaaS, DaaS, IaaS, NaaS, and RaaS.
So now that we understand the Cloud, kinda, manufacturers have introduced something new — very small sensors which can equally communicate and intercommunicate in such a way, justifying a new name, The Internet of Things (IoT), or The Internet of Everything (IoE).
Ostensibly, these little sensors and devices communicate with the Cloud in a two-way format, providing data and receiving instruction. An example of these devices under IoT include sensors designed to control lights and blinds, HVAC systems and appliances, security and energy efficiency systems. More recent additions of IoT devices include wearable medical technologies, wildlife movement monitoring, urban infrastructure monitoring (road and bridge), and even intelligent collision-avoidance sensors in automobiles (both with driver and without driver).
Back to the Cloud. Servers located remotely (in the Cloud) can, and do, communicate with IoT devices out on the edge of the network; centralized computing works for IoT devices. However, propagation delay (another ‘old’ term) has become a serious factor. Propagation delay is the length of time it takes to get a signal from a sender to a receiver and back. Under normal circumstances, while we are impacted by this delay, we don’t really experience it because of our reference point.
Here’s an example. You call a friend who you are meeting at a restaurant, you ask where they are, and then you see them walking around the corner. You see their mouth move, then you hear their voice in your phone. We always have this delay, but our reference point is such that we do not realize it, so it does not bother us.
Not so for IoT devices. These devices need to instantly communicate and intercommunicate between other IoT devices, and the process of these devices speaking to each other in the Cloud, while technically capable, adds way too much propagation delay to the mix. They become ineffective.
This brings a new (old) concept back into play – de-centralized computing. Ahh, remember that? But we can’t call it de-centralized computing because it’s an old term that we were told does not work any longer, so for IoT to IoT device communication a new name had to be created. That name is … The Fog.
And yes, it makes sense. A fog is a cloud at ground level. A billion droplets of water vapor floating around at a low level, not relying on the cloud for existence. And that’s what the idea of intercommunicating IoT devices is. A billion little sensors bouncing around, intercommunicating, and not relying on the Cloud to perform that communication.
In healthcare, IoT is already here and located within wearable technologies monitoring biometric data, in the RFID systems used to track supplies and locate staff, and in mechanical controls for building automation. For hospitals, growth of wearable tech will be seen as the next step, and this growth will be the first impact on architecture from IoT.
Already we are seeing program space being set aside by hospitals to blend clinical engineering, clinical care providers, and IT departments who will work together to choose, fit, configure, and remotely monitor patients wearing sensors, smart clothing, even implants and prosthetics that will communicate back into the hospital network.
While large leaps into IoT and Fog Computing won’t be seen in the typical hospital for a few years, forthcoming IoT devices will route alarms from equipment to care providers, warn of fall risks, automate re-supply of equipment and meds, track clinical process flow, mitigate queuing, and heighten the use of autonomous robots for specimen collection, supply delivery, and remote telemedicine visits. Beyond that, as driverless cars make their way into mainstream, hospital garages and way finding systems will ultimately communicate directly with these vehicles, perhaps even routing cars to appropriate entry points based on the current biometric readings of the passengers within.
The possibilities are, well, still foggy.