On data science and IoT

Since several days have been reflecting on the deep connection between data science (or what we refer to predictive modeling) and IoT. IoT is commonly defined as (Rouse, 2019):

a network of interconnected computing devices, mechanical actuators and sensors able to exchange data between themselves without the need of human interaction.

It’s clear to me that the connection between this new network of things and data science is striking; In fact, I strongly believe the real revolution will come when these two branches of technology will finally be recognized as deeply related. I imagine a future where the data collected from the sensors will be transformed into insight and information by the machine learning algorithms and will automatically trigger a response in the physical world thanks to the physical actuators always connected on the Internet. 

Until now, data science has mostly focused on social network-generated data or Internet generated data (e.g. pictures, text mining on Twitter, etc); the insights that can be gathered from this kind of data is indeed limited in scope because no physical reaction can be triggered; or better, no improvement in efficiency can be triggered by using such data. On the contrary the data generated by the IoT world will pertain to the physical realm: think for example at the footfall in the city or the numbers of/the type of nutrient required by a crop field. All this data will be transmitted automatically and instantaneously over the Internet to algorithms able to predict and decide what to do based. This in turn will trigger a mechanical or chemical action inducing a response that is predetermined by humans using Machine Learning.

It is clear to me that the connection of the two technologies will be very important for humanity at large and it will be a multiplier of human capabilities in almost all fields of the physical realm.