Data analytics has come a long way. Though there is still a long journey ahead, we can now see the road. In this video blog, Big Data Analytics expert Manny Aparicio of Saffron Technology explains why the old approaches to standardisation and business intelligence no longer work. Instead we need data analytics tools that work more like our own human minds.
In the early twentieth century, Henry Ford defined modern manufacturing systems with the production line concept.
Labor was specialized with standardised operation times, operational research was founded as a discipline and the ‘time and motion’ man was feared. The foundations of data analytics were laid. Charlie Chaplin’s ‘Modern Times’ memorably encapsulated the treadmill-like nature of factory work.
With the advent of computers, MRP rose to power then had its loop closed and became ERP, with marketing and human resources inputs. Enterprise efficiencies made a quantum leap upward. Later in the century with emerging ‘fuzzy’ decision support and robotic systems combined with refined quality concepts, standardisation had reached new peaks.
Everything was average, to a high degree, but there were still exceptions. Leading manufacturing thinkers recognise the importance of exceptions, and the opportunity that randomness can create. That’s how penicillin was invented. It is the exceptions that drive us to strive, to research and to innovate.
Much of life is standardised – we share 80% of our genetic code with chimpanzees, but the exceptions matter – they make us what we are. Artificial intelligence has similar concepts. We can model much of daily life, but taking heuristics to the next level is challenging because we have to have the details of the exceptional data if we are to model at the edges. Our models break at the edges, our rules are too rigid and cannot embrace the exceptions.
The latest research tells us that the linear processing of computers (and even the parallel processing paradigm) is not how our brains work. Capturing those data exceptions and turning them into information using advanced data analytics to create linkages in the way that the one hundred billion cells in our brains do, holds promise. It promises to enable us to model real events much more closely. Whether of course that process resembles anything like what happens in the human brain, we shall not know for a long time.
We could see the brain as a black box, and as long as outputs of the data analytics resemble those of the brain, for a given set of exceptional data, then we are on the right track.