he ultimate goal of digitization processes is prediction. This is the greatest competitive advantage that artificial intelligence can bring to an industry or business.
AI performs tasks that are not easy for the human brain: processing large volumes of data, with no apparent connection, in parallel and continuously. The challenge is not to replace human trades with robots and algorithms. The real impact occurs when artificial intelligence is harnessed to do things that already outperform humans today.
When a company does not have access to the analysis of a massive volume of data, it finds itself in a situation where, on the surface, the market suddenly changes them, advertising campaigns have unpredictable results, competitors have jumps in productivity that do not seem have an explanation and in general, they begin to feel that they are in a more volatile and strange situation than the one they knew and mastered until just a moment ago.
Digitization is deceptively complex. An analogy that can be used to better measure the type of project in question would be that of a refinery: data instead of oil.
Before starting, it is required to license a process, an initial know-how, prepared by companies that are experts in the matter and that have been developing and perfecting it for years. Then, it is required to coordinate different disciplines, from chemical engineering (the process), through civil, mechanical, electrical, instrumentation and control engineering. It is also required to evolve from the budget to basic and detailed engineering, purchasing, logistics, assembly, start-up and commissioning. Each of these activities has its analog in a digitization project, although they are more familiar to us in the petrochemical project because we have been doing them for more than a hundred years.
In a refinery, about 70% of the cost and time is spent on piping. In other words, it is the connections between teams that define the complexity and, in the end, the financial and technical success of a project. The same thing happens with digitization: in order to carry it out successfully, it is necessary to know and measure the scope of the task.
The other similarity is that a refinery cannot be built in parts: today let’s put together the distillation column, the furnace and maybe next year we’ll put in the cooling tower. It does not work like that. Each piece is connected and dependent on the others and only when the installation is finished will we begin to obtain the final product. And after a careful process of startups, tests and adjustments of each process unit.
In the world of the twentieth century, the success of economies depended on the number of complex petrochemical projects they could execute. Japan, without a drop of oil of its own, has a petrochemical industry that is still relevant worldwide today. Data is better distributed than oil, but success is reserved for those who can best process it.
