Chapter 2. Studying the IoT Problem Universe and Designing a Use Case

IoT is spread across the length and breadth of the industry. It has touched every possible industry vertical and horizontal. From consumer electronics, automobiles, aviation, energy, oil and gas, manufacturing, banking, and so on, almost every industry is benefiting from IoT. Problems arise in each of these individual business areas that need to be solved connoting the industry it is addressing, and therefore people often segregate the wide spectrum of IoT into smaller and similar groups. Thus, we see names such as Industrial IoT, Consumer IoT, and so on being referenced quite often these days. Keeping aside these broad divisions, we can simply divide the problems to solve in IoT into two simple categories, that is, 'Connected Operations' and 'Connected Assets'.

In this chapter, we will study about the IoT problem universe and learn to design a business use case by building a blueprint for the problem using the problem...

Connected assets & connected operations

With the swift progress of IoT in every dimension in the industry, the associated problems also diversified into the respective domains. To simplify problems, industry leaders took the most intuitive step by defining logical segregations in the IoT domain. Today, there is a plethora of articles and papers published over the Internet, which cite different names and classifications for IoT. As of now, we don't have any universally accepted classification for IoT, but we do see different names such as Consumer IoT, Industrial IoT, Healthcare IoT, and so on. All the IoT-related problems and solutions in the industrial domain were termed as Industrial IoT and so on.

Before studying Connected Assets and Connected Operations, let's explore a simplified classification for the IoT domain. This is definitely not the most exhaustive and widely recognized one, but it will definitely help us understand the nature of the problem better:

When we look at the entire...

Defining the business use case

So far, we have explored what kind of problems arise in a typical IoT scenario and how they can be classified into Connected Operations and Connected Assets. Let's now focus on designing and solving a practical business use case for IoT. We will explore how we can solve problems using the interdisciplinary approach of decision science in IoT.

We'll start with a simple problem in the manufacturing industry. Assume that there is a large multinational consumer goods company, say, Procter & Gamble, who owns a plethora of products. Consider their detergent product, Tide, to study our example. Tide is a detergent powder that comes in liquid form as well, has a variety of scents, different cleanliness levels, and so on. Assume that the company owns a plant in which one production line (the assembly line in which the goods are manufactured end to end) manufactures detergent powder. It manufactures 500 Kgs of detergent powder in a single go. The operations head of...

Sensing the associated latent problems

Problems in real life are often never solo; they are mostly interconnected with multiple other problems. Decision science is also no exception to this feature. While solving a decision science problem, we would often reach a point where we understand that solving the associated problem is more important than the current problem. In some cases, solving associated problems becomes inevitable in order to move ahead. In such cases, we would not be able to practically solve the current problem until and unless we solve the associated problems.

Let's take an example to understand this better. Consider that while solving the problem to identify the reasons for bad-quality detergent manufactured, we inferred that the vital cause for the problem is the difference in raw materials from different vendors or because of insufficient labor in the manufacturing plant (assume). In some cases, the machinery downtime or inefficiency can also be vital reasons for the problem...

Designing the heuristic driven hypotheses matrix (HDH)

Designing the framework for heuristics-driven and data-driven hypotheses forms the foundation of the problem solving framework. The entire blueprint of the problem and problem universe can be captured in this single framework. This isn't a fancy document or any complicated tool. It's just a simple and straightforward way to structure and represent the problem solving approach.

There are three parts to it:

• Heuristics-driven Hypotheses Matrix (HDH)

• Data-driven Hypotheses Matrix (DDH)

• The convergence of HDH and DDH

The heuristics-driven hypotheses is the final and refined version of the hypotheses list that we discussed earlier. The matrix captures every minute detail we need from the hypotheses. It helps us prioritize and filter the hypotheses based on data availability and other results. It also helps us gather all our results in one single place and assimilate in order to render a perfect story. Once the entire HDH is populated, the initial...

Summary

In this chapter, you learned about the IoT problem universe by exploring Connected Operations and Connected Assets in detail. You also learned how to design a business use case for IoT using a concrete example to understand the detergent manufacturing problem in detail and design a blueprint for the problem using the problem solving framework.

This was accomplished by designing the SCQ and understanding how to get started with defining the problem holistically. We also studied about identifying the associated and latent problems and finally explored how to design HDH for the problem.

In the next chapter, we will solve a business use case with a dataset using R. All the context and research gathered in this chapter while defining the problem and designing it will be used to solve the use case step by step.