I have an agenda with this blog: To change the way people use IoT (Internet of Things) and data gathered from it. This is the right time to change how we perceive things when enterprises, companies (small or big) and individuals are planning to invest in IoT and if time and money permits in AoT (Analytics of Things). For more than a decade, we have seen how web analytics has evolved itself. For a bigger part of this piece, I have used analogy of web analytics to make it more simple for a novice to understand the concepts of AoT. So as was applicable on world wide web in early years, we have been relying on faith-based initiatives when it comes to IoT, its Data collection, process and Analysis. Which is obvious because that is how we decide in our daily life and professional careers as well. But do we really need to do it when we have enough data to bank upon?
Internet of Things, according to me are “Devices which can generate and share/ communicate data over Internet with other devices”. So essentially , the basic infrastructure of an IoT system are
- Devices (Things able to capture information)
- The Network (Medium to communicate)
- System (to process information generated by things).
At present, we live in the most data rich eco-system in the history of mankind. We are on the quest of finding such other lives on distant planets with the help of this data ecosystem. Usage of data is becoming more popular day by day as we are getting used to smart devices, internet and connectivity. Its easier to communicate, travel, work, run or even sleep with the help of data. Today we want our surroundings to be smarter than yesterday. We want our needs to be understood and this more or less comes out as a pattern it starts watching closely. Artificial intelligence is an advance science of these patterns only. This blog will talk about things which surround us with lesser degree of complexity or at least what they appear to be on surface. I will let you decide as at times simplest things are most complex to design. I might be digressing but the point what I am trying to reach is that these patterns are recordable and hence create a data bank. If you can design a system which can read the patterns hidden within the data, this can help you to unfold the true potential of Analytics of things; you will begin to decipher the hours of readings, numbers, patterns etc.
Now put this in the business scenarios. We can use the data to determine how my machines are working, predict breakdowns, Invest in lesser inventories, lesser WIP, Shorter Turn Around Times (TAT), higher machine performance, Low insurance premiums and what not. Now in an individual case one can take Prescriptive Decisions to invest in his own health and time, avoid accidents in households, low health insurance premiums and the applications is limitless. If you know how to use the data right, you are the next billionaire and on the cover of a Tech Magazine. Many of them already are. But this blog is not about how to become a billionaire (if you know do let me know too), rather how to exercise right usage of IoT data. AoT is not about simply collecting the data, data processing and simple insights more like a salsa between your heart and your brain. It is Art and science applied simultaneously. It will tell you what is the economic value of the data generated by IoT in daily routine if you go through rigorous resultant analysis in the right direction. Imagine the below situation for a better understanding (Refer Fig 1). Being from the digital marketing world I worked upon huge chunk of web generated data, yet critical at times. But the lack of data types and depth leads to very less of meaningful data. As said by Avinash Kaushik , “I say this only partly in jest, we could blame incompetence on the lack of sufficient types of data.” So we always have a get-out-of-jail-free card.
I would like to introduce Trinity of AoT (credit to a best seller on Web Analytics called Web analytics: An Hour A day) here applicable for Analytics of Things. Trinity of AoT i.e Experience, Behaviour and Outcomes. Desired Experience-> Read Behaviour -> Correction in Outcomes (Refer Fig 2). Why “Trinity” because these three processes need others to exist for its own existence. I have modified to put a sequence in IoT context where I have taken Bottom to top approach. You set a target/task assigned to system to reach a desired experience. So you start noticing behaviour because deviations occurs. But reading behaviour will tell you What but not why ? We have the what : what part of machine is not working as desired? What is the average time taken to complete the task ? What sources are causing deviations ? what is and what is not ? But not Why. So You start noticing Outcomes and put a feedback in the system to improve results and decrease the gap in every iteration till we reach desired experience. Please note this is applicable only on certain IoT system which is targeted to reach such desired experience only. Any Auto Tuning system (with a feedback) could be an example.
I would like to stop here and wait for the feedback if you have read this far. I will write more IoT and its possible applications in coming weeks. I would like to know what else you want to know about IoT and AoT. Help me to help you better.