In this special guest feature, Brendan O’Brien, Co-founder and Chief Evangelist at Aria Systems, discusses the critical role machine learning has to play in the all-important “Analyze” phase of IoT services. Brendan introduced cloud billing before the concept of “cloud” was even on the horizon. The number-one ranked cloud-billing provider, Aria, is at the forefront of the recurring revenue revolution, empowering enterprises to monetize IoT and grow recurring revenue at scale.
At its core, the Internet of Things is about sensors embedded into devices of all kinds, which provide streams of data via internet connectivity to one or more central locations. The purposes for transmitting sensor data are myriad, but the assumption in all cases is that that data can then be analyzed and acted upon in some way that is beneficial to the user. In short, this means that all IoT-related services, no matter how disparate they may be, always follow these five basic steps:
- Sense
- Transmit
- Store
- Analyze
- Act
Origins of IoT
If I ask myself how the IoT came to be, the shortest answer I can provide is that good ‘ol Moore’s Law made the first three steps in this chain (Sense, Transmit, and Store) ubiquitous and commoditizable. The hardware, software, and connectivity required to perform these steps has become very small, very cheap, very efficient, and very broadly available. When we hit the point of critical mass a few years back when all of those “verys” became applicable qualifiers, the IoT was born. It’s critical to note here that these first three steps were subject to this commoditization because, essentially, they are relatively “dumb”: long-existent capabilities coupled with standard coding make them all possible.
Demonstrating Value
However, for any IoT application to be worth buying (or making), it must demonstrate value in the last step of that chain, the “Act”. Of course, “act” can mean an infinite number of things, ranging from a profound physical action (e.g. deploying an ambulance to the site of an auto accident) to merely providing basic information to a relevant consumer (e.g. sending a text message to alert a driver that their car needs an oil change). But no matter what the ultimate step of “Act” actually is, its worth is entirely dependent on the penultimate step that preceded it: “Analyze”.
Where Machine Learning Comes In
It is here, at the “Analyze” step, that the true value of any IoT service is determined, and this is where artificial intelligence (or, more properly, the subset of AI called “machine learning”) will provide a crucial role.
Machine learning is a form of programming that empowers a software “agent” with the ability to detect patterns in the data presented to it so it can learn from these patterns in order to adjust the ways in which it then analyzes that data. We already experience benefit from machine learning in our everyday lives when Netflix gives us a tailored movie recommendation or Spotify modifies our playlist. When machine learning is applied to the “Analyze” step, it can dramatically change what is (or is not) done at the subsequent “Act” step, which in turn dictates whether the action has high, low, or no value to the consumer.
Does IoT Really Need AI?
Now that we’ve discussed the critical role Machine Learning has to play in the all-important “Analyze” phase of IoT services, we can expand upon that point by examining simple hypothetical versions of a service both with and without the benefits of machine learning.
IoT Without Machine Learning
Imagine a wearable device, designed to track vital signs in order to alert the wearer to dangerous health conditions. If the device senses a highly elevated heart rate, its job is to recognize that fact and act on it in some valuable way, which might include automatically summoning an ambulance. But an elevated heart rate alone isn’t necessarily indicative of a serious problem: it could be the result of a serious medical problem, or it could just be due to a strenuous workout. Without machine learning that adjusts its analysis (or heuristics) based on patterns about the wearer that have been divined over time, there’s no way to tell the difference, with massively disparate consequences to the consumer of the service.
IoT With Machine Learning
While many digital services don’t yet take advantage of the massive benefits Machine Learning can offer to their users, the great news is that some already do. Companies like Amazon and Netflix are busy utilizing machine learning in their offerings in order to more intelligently promote targeted content to users, going far beyond merely “customers who bought what you bought also bought X”. There’s still a long way to go, however.
If we take the analogy of the wearable one step further and ask ourselves what the ideal level of value a service like that should be, we’re immediately looking at a combination of decisions and actors (ambulance companies, insurance companies, hospitals, primary care physicians, etc.) that could and should play a critical role in the outcome of a drama that began with a simple, commoditized sensor in an inexpensive device. Only sophisticated machine learning is capable of providing the means to orchestrate these decisions, and making the Internet of Things all we dream it can be.
Sign up for the free insideBIGDATA newsletter.
Speak Your Mind