If the gap between what we already know and what is new is too wide, if the new input we receive is too distant from what we are already familiar with, there can be no learning.
This is one of the reasons why true innovation requires time. It first needs to be digested by a small group of early adopters, technological pioneers, individuals who are willing to challenge the conventional way in which things are done. When they succeed, when others start seeing that after all the new path is not only safe, but it's also a more sensible choice, then you have disruption. Then you change the rules of the game so radically that it becomes difficult to remember how things were before the new normal.
This is what happened in all human enterprises from the very beginning of our common history. After all, change is first a matter of awareness that something is not right. Awareness leads to determination which leads to action.
There is no doubt now that a new era of digitalization is upon us. As machine learning improves, more and more processes can be effectively automated, and many tasks can be optimized. Low-IQ digital solutions used to do little more than their analogue predecessors. The fuel gauge, for instance, would let us know how much gas we have in our tank.
Now computers can provide us with a relatively accurate projection of how many miles we can cover based on our driving history. Computers can also calculate the most effective route based on several parameters, making the transportation of people and goods an increasingly safer, more efficient, and therefore more sustainable affair.
And then comes data fusion. More than a software solution, data fusion is a way of understanding data based on quantity and, crucially, quality. Data fusion architectures like Navidium's do more than getting huge amounts of data and organizing them following some basic, pre-determined rules. Data fusion is the answer to the challenges posed by the digital era: real-time responsiveness, ease of use and interaction between human and machine, and a truly stochastic approach to understanding the world.
Responsiveness is required everywhere as the world accelerates and the frequency with which we act increases. Computers can significantly decrease the time it takes to respond to events and dramatically increase operational efficiency when everything goes according to plans. But variability happens and predictions are broken. That's when, in the words of Shyam Sankar (COO at Palantir), the role of human-computer cooperation becomes apparent.
For this cooperation to be fruitful and beneficial we need excellent human-computer interfaces, architectural designs that make interactions smooth, seamless, and as effective as they can be. But, again, variability happens. Different types of interaction arise, requiring different types of interfaces or, when possible, a single dynamic and adaptive interface, capable of learning new forms of interaction.
The ultimate challenge is therefore a matter of learning beyond the illusion of entirely controlled settings. Instead, human-machine learning must be based on a stochastic approach, one in which unpredictability is quantified in terms of probability, not just vague possibilities.
This is the core of Navidium's data revolution: our software is designed to learn to recognize patterns in complex, random, sequences of events, estimating when one route will make you save time, fuel, or both, based on accurate weather forecasts, environmental data, and, crucially, on the actual condition of the ship. This intelligence is then presented on board and ashore in a clear and accessible way, making data truly actionable and adding human expertise to the equation.
In conclusion, we are living a digital revolution, one that is likely to lead to a safer, greener, and more sustainable future in which the power of data will be made dynamically accessible to many, perhaps all. This is just the beginning.