Machine-generated Data vs Human-generated data
Data is the new resource in the 21st century, and it has become as important as ever, a lever for any businesses that can turn a business problem into a data problem. Human-generated data has existed for a while.
Back when the ancient city builder decided on how many wells or canals to dig, it was largely dependent on the population that lived within that area. Back then, it was largely drawn from the experience and tradition of the city builder to decide what and where things are. It works because experience helps them avoid pitfalls that can be costly, and traditions help to bring a sense of familiarity of how the design of the well or canal should be, however little value these designs provide.
The reliance on experience works because it helps users draw from the past consciousness or subconsciousness to navigate the present. However, Inconsistency ensured.
Get Darius’s stories in your inbox
Join Medium for free to get updates from this writer.
Machine-generated data is new to the scene. Back during the early days of machine maintenance, monitoring the internal state of machine is more of a visual approach. Looking at the cogs and wheel to check if there is something out of the ordinary.
As machine design gets more complicated, machine designers introduce sensors to monitor states of the machine at different stages. Temperature, number of oscillations per second, etc.
However, there is always an aspect of human intervention because sensors can get faulty, a situation of who checks the checker. Also, you cannot put sensors on every type of machine fault that we identified. The cost outweighs the advantage.
Machine-generated data (what temperature is this transformer?) and human-generated data (Is there a crack on the transformer?) have to coexist together for now.
Humans having the final say in the maintenance action is to have accountability. Without that, who is to be responsible when the machine is at fault?
