In this blog post, you can learn how AI principles are adopted while developing configurator models in SAP PLM VC
Before we start, first let us understand some basic things about Artificial Intelligence
Artificial intelligence (AI) is the ability of a computer program or a machine to think and learn. It is a simulation of human intelligence processes by machines, especially computer systems. These processes include learning (the acquisition of information and rules for using the information), reasoning (using rules to reach approximate or definite conclusions) and self-correction
AI is concerned with below two basic ideas:
The collection of knowledge related to a problem (or an opportunity) to be used in an AI system is called as Knowledge Base.
AI has below two basic elements:
Knowledge Pyramid also known as DIKW hierarchy (Data-Information-Knowledge-Wisdom) in AI is an often-used method, with roots in knowledge management, to explain the ways we move from data (the D) to information (I), knowledge (K) and wisdom (W) with a component of actions and decisions. It refers loosely to a class of models for representing purported structural and/or functional relationships between data, information, knowledge and wisdom.
A traditional data-information-knowledge-wisdom pyramid is pictorially represented as below
Product pyramid in case of CTO scenario can be represented as below
The business-driven rules and the engineering-driven rules need to be integrated within a methodology and main objective is that its look and feel needs to meet the company’s business goals.
Complex products pose true configuration problems. The presentation should make it as easy as possible for the user, but the underlying engine must be able to express complex relationships and apply them in all possible combinations
A complex product might turn into a lot of material numbers. The business process determines what the sales configurator outputs
The Knowledge Base is focused on presenting the product at sales time to the customer, which can increase sales, but
SAM PLM Variant Configurator is not just a computer application; its origin is in AI; it encapsulates an array of systems, all of which, together, replicate the processes of thought.
Developing Configurator models for CTO(Configured to Order) and MTO+E (Make to order with Engineering) scenarios is the processes of thought for problem solving: Abstraction / Analysis / Deductive Inference
First Principles – the AI (Artificial Intelligence) Heritage
First-principles thinking is one of the best ways to reverse-engineer complicated problems and unleash creative possibility. Sometimes called “reasoning from first principles,” the idea is to break down complicated problems into basic elements and then reassemble them from the ground up.
Conceptual Framework Applied – “in practice”
Thus, for example…
Declarative Reasoning – talking about “real things” (Dependencies)
What are the “laws” or principles underlying its behavior?
Can these “laws” or principles be stated mathematically?
Can they be stated “physically”?When we say “why,” are we describing real behavior?
Can they be stated as constraints? Are they always true?
Hence we can conclude by looking the above example of features of table (product), we need to consider and adopt AI first principles while developing a configurator model.
External References:
Knowledge Pyramid:
https://www.i-scoop.eu/big-data-action-value-context/dikw-model/
Expert system:
http://intelligence.worldofcomputing.net/ai-branches/expert-systems.html#.XRV6BogzbIU
https://www.tutorialspoint.com/artificial_intelligence/artificial_intelligence_expert_systems.htm
knowledge base:
https://searchcio.techtarget.com/definition/knowledge-based-systems-KBS
First principle:
https://fs.blog/2018/04/first-principles/
Before we start, first let us understand some basic things about Artificial Intelligence
Artificial intelligence (AI) is the ability of a computer program or a machine to think and learn. It is a simulation of human intelligence processes by machines, especially computer systems. These processes include learning (the acquisition of information and rules for using the information), reasoning (using rules to reach approximate or definite conclusions) and self-correction
AI is concerned with below two basic ideas:
- involves studying the thought process of humans.
- deals with representing those processes via machines
The collection of knowledge related to a problem (or an opportunity) to be used in an AI system is called as Knowledge Base.
AI has below two basic elements:
- Fact (or data) about a problem domain.
- Rules that direct the use of data to solve specific problems in the domain
Knowledge Pyramid also known as DIKW hierarchy (Data-Information-Knowledge-Wisdom) in AI is an often-used method, with roots in knowledge management, to explain the ways we move from data (the D) to information (I), knowledge (K) and wisdom (W) with a component of actions and decisions. It refers loosely to a class of models for representing purported structural and/or functional relationships between data, information, knowledge and wisdom.
A traditional data-information-knowledge-wisdom pyramid is pictorially represented as below
- Data – refers to numeric strings that by themselves do not have a meaning. Can be facts or figures to be processed.
- Information – is data organized so that it is meaningful to the person receiving it.
- Knowledge – is a clear and certain perception of something or organized information applicable to problem solving.
- Wisdom – Knowledge that has been experienced
Now we can see how configurator model’s development process in SAP PLM VC can be mapped into AI Principles
Product pyramid in case of CTO scenario can be represented as below
The business-driven rules and the engineering-driven rules need to be integrated within a methodology and main objective is that its look and feel needs to meet the company’s business goals.
Complex products pose true configuration problems. The presentation should make it as easy as possible for the user, but the underlying engine must be able to express complex relationships and apply them in all possible combinations
A complex product might turn into a lot of material numbers. The business process determines what the sales configurator outputs
- a single material, with attribute values, to be further expanded before production.
- a list of sales materials, to be further expanded before production.
- A full list of production materials, ready to go to manufacturing.
The Knowledge Base is focused on presenting the product at sales time to the customer, which can increase sales, but
- It is even more powerfully aimed at
- Increasing customer satisfaction (chooses the right thing)
- Decreasing the costs of taking sales orders (fewer mistakes)
- Decreasing the cost of production (fewer products, more variance, variance in a very well-specified format.
- The knowledge base as part of a rigorously conceived business process.
SAM PLM Variant Configurator is not just a computer application; its origin is in AI; it encapsulates an array of systems, all of which, together, replicate the processes of thought.
Developing Configurator models for CTO(Configured to Order) and MTO+E (Make to order with Engineering) scenarios is the processes of thought for problem solving: Abstraction / Analysis / Deductive Inference
First Principles – the AI (Artificial Intelligence) Heritage
First-principles thinking is one of the best ways to reverse-engineer complicated problems and unleash creative possibility. Sometimes called “reasoning from first principles,” the idea is to break down complicated problems into basic elements and then reassemble them from the ground up.
Conceptual Framework Described – “the theory”
- Abstraction = Generalization = Classification
- The formulation of generalized concepts by extracting common qualities from specific examples; the formation of general principles from detailed facts
- Applying the principles of Generalization, Decomposition, and Logical Dependency…
- Rigorously, formally, and with discipline
- Analysis = Decomposition
- The division or decomposition of a physical or abstract whole into its constituent parts to examine or determine their relationships
- Describing a product as an object, which can be…
- Abstracted (I.e., generalized, abstracted, classified and characterized)
- Analysed (I.e., broken down into its constituent objects, decomposed)
- Inferenced about (I.e., given logical relationships and dependencies)
- Deductive Inference = Logical Dependency
- The process of reasoning by which a specific conclusion necessarily follows from or depends upon a set of specific premises
- Or, from a slightly different perspective…
- To describe a product, or a set of products, as an object, or a set of objects, with properties (characteristics) and behaviors (rules of dependency – logical inference)
Conceptual Framework Applied – “in practice”
A Product Modeling Methodology – based on First Principles
- Object-orientation – talking about “real things” (Classification)
- What is a “product object”?
- What it is not…
- A part number, or, in most cases, a “product number”
- Absolutely concrete, something you can touch and feel
- What it is
- Something that corresponds to a picture you have in your mind
- Concrete, because it corresponds to a “picture”
- Abstract, because the “picture” exists only in your mind
- What it could be…
- A panelboard, a breaker, a switchboard, an MCC
- A table, a cabinet, a car, a diesel generator, a truck, a plane
- A computer, a PBX, a GSM Network, a network server
- What is a property / characteristic of a “product object”?
- While there are no such “things” as…
- Height, Color, Depth, and Weight;
- Trees can be “tall,” Leaves can be “green,” Holes can be “deep,” and Stones can be “heavy”
Or stated slightly differently, and in regard to a specific Tree, a specific Leaf, a specific Hole, and a specific Stone…
- height = “tall”
- color = “green”
- depth = “deep”
- weight = “heavy”
Object-orientation – talking about “real things” (Decomposition)
- What is a “component” of a “product object”?
- A “component” can be imagined, and it can be named, and it has a reality of its own
- It is part of something else
- It has its own existence, and can exist by itself
- It has its own characteristics, which are not only proper to it, but cannot and should not be confused with the characteristics of its “parent”
- A “component” can be imagined, and it can be named, and it has a reality of its own
Thus, for example…
- The Legs of a Table are “tall” – their “tallness,” however, cannot be confused with the “tallness” of the Table
- Table-Slab.Thickness = Table.Height – Table-Leg.Height
Declarative Reasoning – talking about “real things” (Dependencies)
- How does the “product object” act? (Objects and their Behaviors)
What are the “laws” or principles underlying its behavior?
- The “height” of a Table’s Legs is less than the “height” of the Table itself, but do we know why?
Can these “laws” or principles be stated mathematically?
- The “height” of a Table’s Legs is less than the “height” of the Table itself, but can we say why?
Can they be stated “physically”?When we say “why,” are we describing real behavior?
Can they be stated as constraints? Are they always true?
- 5x * 6x = 30x; 30x / 6x = 5x; 6x = 30x / 5x
- Table-Slab.Thickness = Table.Height – Table-Leg.Height”
Hence we can conclude by looking the above example of features of table (product), we need to consider and adopt AI first principles while developing a configurator model.
External References:
Knowledge Pyramid:
https://www.i-scoop.eu/big-data-action-value-context/dikw-model/
Expert system:
http://intelligence.worldofcomputing.net/ai-branches/expert-systems.html#.XRV6BogzbIU
https://www.tutorialspoint.com/artificial_intelligence/artificial_intelligence_expert_systems.htm
knowledge base:
https://searchcio.techtarget.com/definition/knowledge-based-systems-KBS
First principle:
https://fs.blog/2018/04/first-principles/
- SAP Managed Tags:
4 Comments
