How to Design Decision Making? : For Social Innovation with Numerical Optimization (Part 2)
In this column, Mr. Hirai from Hitachi Consulting, who supports the dissemination of Annealing Cloud Web, will introduce the methodology for implementing numerical optimization technology in society.
This column is filled with hints for advancing issue clarification and requirement definition, so if you're unsure where to start or how to articulate your thoughts as you progress in your study of numerical optimization, please take a moment to read it.
Nobuyuki Hirai
Joined Hitachi Consulting in 2017. Engaged in data analytics projects such as proof of concept and requirement definition for social application of machine learning and numerical optimization. Utilizing his knowledge and experience, he is also involved in activities such as internal education on the application and utilization of numerical optimization technology, as well as creating educational content for the Hitachi Group.
He also applies his experience in studying design during his university days to UI/UX design of optimization systems and design study sessions within the company.
Introduction
In this column, I would like to introduce the know-how for applying numerical optimization technology, including CMOS annealing, to issues that require human decision-making.
In the first part, I introduced the method of issue identification using the decision architecture design framework (FW) that we use, through familiar decision-making scenarios as examples.
In this part, I would like to introduce some tips for using the FW, the commonalities between familiar issues and business issues, and hints for discovering business use cases for optimization technology.
Summary
Tips for using the decision architecture design FW
I would like to introduce three tips for visualizing issues using the decision architecture design FW that I explained in the previous column.
The first tip is to "The skeleton of the issue should be defined first, and the outline should be defined afterward".
The above tips are not mandatory, but in my experience, it is more effective to first define the skeleton of the issue, such as value and action. Then, define the outline of the issue, such as rules, preconditions, reference information, and data. This approach leads to a clearer visualization of the issue.
The second tip is to "Draw the whole picture of the issue using the power of common sense".
The rule explained in the first part, "you cannot cook the noodles without first boiling the water," may seem obvious. However, if you do not understand that "boiling" is a cooking method that involves putting ingredients in hot water, you cannot identify this rule. There are many implicit rules and assumptions in the human mind. If these are not explicitly stated, the quality of the issue, and consequently the quality of the solution, will be incomplete.
Therefore, as shown in Figure 2-1, using the power of common sense to identify the items to be listed in the FW one by one is an effective approach for visualizing the issue.
Click to enlarge the image.
The third tip is to "Divide the issue".
For example, when considering the issue of SCM (Supply Chain Management) plan optimization, attempting to optimize the entire system at once often results in a problem size too large to be solved with current computer technology. However, within SCM planning, there are many sub-issues (see Figure 2-2). By breaking down these sub-issues into individual task levels, you can make the issues manageable in size and solvable in reality. Additionally, if you can define the relationships between sub-issues as rules, it is possible to define and solve multiple issues as a single issue.
Click to enlarge the image.
The words left by René Descartes, "Divide each of the difficulties under examination into as many parts as possible, and as might be necessary for its adequate solution," hold true even in the social implementation of optimization technology.
Commonalities between familiar issues and business issues
In this column, I have been discussing familiar issues as examples, but here I would like to discuss the commonalities between these issues and business issues.
First, let's consider the example of ingredient procurement. This takes the same form as the issue of optimizing vehicle delivery schedules in logistics, as shown in Figure 2-3a. Such issues, where multiple locations are visited in a single continuous path, are known as the traveling salesman problem (TSP) in the world of numerical optimization and are considered one of the typical problems in this field.
Additionally, the cooking example of Spaghetti Aglio e Olio takes the same form as scheduling plans like production planning (see Figure 2-3b). By replacing the stove and cutting board with manufacturing equipment and each task with orders processed by each piece of equipment, the Spaghetti Aglio e Olio cooking optimization model transforms into a production planning optimization model that enhances time efficiency. Such problems are known as job shop scheduling problems, which are also considered one of the typical problems in the field of mathematical optimization.
Click to enlarge the image.
The forms of the two issues are shown in Figure 2-4. By focusing on the actions of each issue, you can see that the actions for ingredient procurement and delivery planning, as well as for cooking and production planning, have the same structure. Since there are several types of issues like this, you can deepen your understanding of numerical optimization through familiar issues and develop the ability to adapt this knowledge to business issues.
Click to enlarge the image.
Human cognition and decision architecture design
Let's apply the four elements of the decision architecture design FW to the human cognition process. Please refer to Figure 2-5. Information and rules from the outside are input into the brain through the five senses and learning. This information is then integrated with the existing information and rules already present in the brain to understand the situation. Decisions are made in accordance with one's values, leading to actions. Seen in this way, we can realize that we are constantly optimizing unconsciously in our lives; that is, Cognition = Optimization.
Click to enlarge the image.
Objectively perceiving one's own cognition is called "metacognition." The social implementation of optimization technology begins with visualizing issues through the decision architecture design FW.
OODA loop: Digital twin and decision architecture design
Figure 2-6 shows the OODA loop, which was devised by John Boyd, a U.S. Air Force officer, with the technologies corresponding to each quadrant of the loop and the four elements of the decision architecture design FW plotted in.
Click to enlarge the image.
This loop, modeled after the thought patterns of fighter pilots in rapidly changing air combat situations, is closely related to concepts such as digital twins and CPS (Cyber-Physical Systems). In other words, the left quadrant represents the visible physical space, while the right quadrant represents the invisible cognitive/cyber space. Data obtained from the physical space is input into the cyber space (Observe), where the situation of the modeled issue is understood, decisions are made (Orient and Decide), and appropriate feedback is given to the physical space (Act).
In the OODA loop, optimization technology is responsible for the Orient and Decide quadrants. Decision architecture design can be described as a method for designing issues in the cognitive space as issues in the cyber space. Modern products and services are built from a combination of various technologies. By combining these technologies to enable the OODA loop, promising use cases for digital twins and CPS can be created.
Afterword
In this column, I discussed tips for using the decision architecture design FW, the commonalities between familiar issues and large-scale business issues, and hints for discovering business use cases for optimization technology.
I hope that you will use the decision architecture design FW to discover and visualize issues as optimization use cases, and take the first step toward the social implementation of optimization technology.
I also plan to continue delivering information related to decision-making and optimization, so I would be grateful if you could occasionally check this site.
Thank you for reading this far, and let's meet again.
