Decision Making Under Uncertainty and Risks in the Face of Rapidly Advancing Technologies

Introduction

Today decision making in science, technology and business is based on measurements and metrics that facilitate or justify the selection of one alternative over others. The current state of the technology allows for monitoring of almost any parameter that may be relevant in a decision. This is achieved by a multitude of connected sensor devices. This chapter is intended to suggest how the understanding of risk requires some new approaches since measurement tools have changed and have been refined with the new and advanced technology.

The concept of “uncertainty” obviously implies a lack of certainty, understanding it as a clear knowledge of something, or the firm adherence of the mind to something knowable, without fear of any error. On the other hand, the term “risk” is often associated with similar characteristics, although it adds negative effects depending on the decision taken, which can be quantified in terms of probability.

“Quantification” precisely implies the fact of numerically expressing a magnitude, so quantifying uncertainty or risk implies the numerical expression of something that is not knowable or not clear. Mathematically, this is equivalent to the complementary value of what is certain and therefore, adding both values results in the whole unit. Other meanings for the concept of “quantification” consist of introducing the principles of quantum mechanics in the study of a physical phenomenon, which would bring here the Heisenberg Uncertainty Principle.

In either case, uncertainty is a fascinating topic that encompasses a wide range of scientific and practical fields. The economy, physics, decision theory, risk assessment are areas of knowledge where this term has traditionally been treated. More recently, the analysis of uncertainty has become common in mathematical modeling, numerical analysis, advanced statistics, information technology, information technology and communication.

For example, in the specific case of an industrial design, this can be considered as an iterative process of reducing uncertainties, conceiving an idea with the intention of materializing it into a product, which must fulfill the required functions during a foreseen period of time. A good design should prevent the occurrence of failures during the life cycle of the product, satisfying the requirements of safety, environmental protection, and compliance with laws and regulations (Sanjurjo, 2013). A risk analysis therefore at this early stage, would identify and correct possible future scenarios of failures.

Just as a risk analysis is applied here for an industrial design, it is equally useful for other areas. In fact, the study of uncertainty and risk has been in contact with issues such as epistemology, administration science, psychology, public debate and the theories of democracy.

On the other hand, Industry 4.0 refers to a fourth generation in the manufacturing and productive activity, consisting of the so-called Fourth Industrial Revolution. This fourth revolution is characterized (Lasi et al. 2014) by intelligent systems based on the Internet of Things (IoT) solutions.

One of the application areas of Industry 4.0 is precisely to reduce the uncertainty in decision-making where, in an autonomous way, the intelligent system predicts errors, performs diagnoses and triggers actions in order to reduce risks at the same time that the process gains operative efficiency. Obviously, these systems demand a lot of data with high quality, coming from relevant information from multiple sources (Lee et al., 2014).

The digital revolution begins at the end of World War II when the transistor was developed, the passage to the integrated circuit was immediate and from that moment onwards, the process capacity has been growing exponentially following the well-known Law of Moore that predicted that every twenty-four months the processing capacity of microprocessors would double. This growth of the processing capacity and the storage of digital data, together with the connectivity provided by the Internet, has brought an unprecedented tractor effect in all fields of science and technology (Sanjurjo, 2016). The great advance, much less visible to the normal citizen, is recently taking place in the field of synthetic or artificial intelligence (AI), although the new techniques of neural networks have meant the definitive takeoff of this technology which is already present in many facets of our lives and, specifically, the reduction of uncertainties when making decisions or selecting alternatives.