The fuzzy logic, different types of AI, and their contributions to Design for Lean and Enterprise 4.0.

By Rémy Rodriguez, Master Black-Belt Lean Six Sigma

1. Introduction

Industry 4.0 represents a profound transformation of industrial processes through the integration of advanced technologies such as the Internet of Things (IoT), artificial intelligence (AI) ?and cyber-physical systems. This framework is based on the interconnectivity of machines, systems and data, thus enabling intelligent and autonomous production processes. At the same time, Lean Management? aims to eliminate waste in production processes while optimizing added value. The combination of these two approaches, called "Design for Lean",? promises to improve the operational efficiency of companies by using advanced technologies.

This article explores how fuzzy logic? and different types of IA can help optimize Design for Leanin a context of Enterprise 4.0. By combining the ability of fuzzy logic to model uncertainties with the power of AI to handle large amounts of data, companies can streamline their processes while gaining flexibility. The objective is to demonstrate how these technologies can meet modern industry challenges and provide more adaptive and robust systems.

?

2. Fuzzy logic: a brief introduction

Fuzzy logic, introduced by Lotfi Zadeh in 1965 , is a mathematical model that allows to deal with uncertainties and inaccuracies, unlike binary logic which works on absolute values of true or false. In a fuzzy system, variables can have intermediate values between 0 and 1, which better reflects often complex and ambiguous real situations. This approach is particularly useful for modelling systems where boundaries between states are blurred, such as quality management, process efficiency or decision-making in uncertain contexts.

In an industrial setting, fuzzy logic has been successfully applied to areas such as</b> robotic system control,</b></b> supply chain optimization, and predictive maintenance. By allowing complex situations with incomplete or ambiguous data to be modelled, it helps improve system performance by addressing uncertainties that traditional methods cannot effectively manage. These capabilities make it a key technology for modern industry, especially in the management of non-linear processes.

?

3. Types of artificial intelligence and their contributions

3.1? AI technologies are classified according to their ability to mimic human characteristics, the technology they use to do so, their real-world applications and the theory of mind, which we will discuss in more detail below.

Using these characteristics as a reference, all AI systems – real and hypothetical – fall into one of the following three types:

• Narrow artificial intelligence (ANI),? which has a narrow range of capabilities;
• General artificial intelligence (AGI),? which is up to human capabilities.
• The artificial superintelligence (ASI),? whose capabilities are superior to those of humans.

3.2? We will focus on two sub-branches of ANI and their combinations through a hybrid AI.

Symbolic AI, which is based on explicit rules and logical algorithms, was one of the first paradigms in the field of artificial intelligence. It is effective in solving well-defined and structured problems such as expert systems. However, its limitations become obvious when it comes to modelling systems that are more vague or imprecise. In the context of Design for Lean, symbolic AI can play a role in automating well-defined decision-making processes, but it may lack flexibility in more dynamic and uncertain environments.

Connectionist AI, especially with artificial neural networks, excels at managing fuzzy data and complex environments. By learning from large amounts of data, this form of AI is able to make predictions and make decisions based on patterns even when the data is inaccurate or incomplete. In Industry 4.0, these neural networks are often combined with fuzzy logic to create "neuro-fuzzy" systems capable of optimizing lean processes by adapting real-time decisions to changing conditions.

Hybrid AI, Hybrid? AI tries to bring together the best of both worlds: explicit knowledge manipulation (symbolic AI) and learning from data (connectionistic AI). By combining these two approaches, hybrid AI allows a system to become both flexible and interpretable, while still being able to handle large amounts of complex data.

For example, in a hybrid system, the logical rules defined by the symbolic AI can guide the neural network in structuring and interpreting data. Conversely, machine learning can refine or adjust expert system rules based on new data. This allows for symbolic reasoning assisted by machine learning. Another example of hybrid AI is the use of a neural network to learn from unstructured data, and then applying a symbolic model to interpret or validate the results.

3.3? Hybrid AI has applications in a variety of fields.

Health: In medical diagnostic systems, hybrid AI can use established clinical rules to interpret the results of medical image-based learning. This both leverages the massive data in health records and ensures a clear explanation of diagnoses.

Industry 4.0 : Predictive maintenance systems can combine the ability of neural networks to detect anomalies from sensor data (connectionist) with symbolic fuzzy rules to make more reliable and accurate maintenance decisions.

Robotics : In cognitive robotics, hybrid AI enables robots to learn from their environment (connectionist) while integrating logical rules to ensure safety or follow specific procedures (symbolic).

?

4. Synergies between fuzzy logic and AI in Design for Lean

Fuzzy logic allows for improved Design for Lean by modelling processes where data is uncertain, such as demand forecasts or machine conditions in a production line. For example, in a lean approach it is often difficult to accurately anticipate the volumes of production needed. Fuzzy logic can be used here to allow companies to model these uncertainties in a more flexible way, so as to minimize waste while responding to fluctuations in demand.

By combining fuzzy logic with AI systems, companies can also improve their ability to react in real time to unforeseen events.? For example, an AI-fuzzy system could continuously analyze production data to automatically adjust processes, reducing interruptions and optimizing resource utilization. This type of application allows to maintain the Lean efficiency while increasing the resilience of industrial processes in the face of disturbances.

?

5. Business 4.0 and fuzzy logic

In Industry 4.0, cyber-physical systems (CPS)? play a crucial role by integrating physical and digital elements to optimize production. Fuzzy logic is particularly useful in these systems, as it can manage the complexity of interactions between machines, sensors and changing environments. By taking into account fuzzy parameters such as machine wear or fluctuations in demand, CPS can adjust processes in real time to maximize performance while minimizing costs.

Another key area is predictive maintenance, where fuzzy logic combined with AI can analyze sensor data to predict future failures</b>. For example, in a 4.0 plant, fuzzy algorithms can continuously evaluate the condition of machines based on ambiguous measurements (such as temperature or vibration) and trigger preventive actions before a breakdown occurs. This not only reduces downtime, but also improves the overall efficiency of production processes.

?

6. Conclusion

Recent advances in AI and fuzzy logic offer important opportunities for process optimization in a Lean approach within Enterprise 4.0. The ability of fuzzy logic to handle uncertain environments, combined with the effectiveness of AI in analyzing massive volumes of data, offers more flexible and adaptive solutions to modern production challenges. By integrating these technologies, companies can not only increase their operational efficiency but also improve their ability to respond quickly to disruptions.

The future of Industry 4.0 lies in intelligent automation and continuous optimization . Hybrid systems combining fuzzy logic and different types of AI promise to go beyond traditional processes, offering significant gains in productivity and resilience.? These advances should profoundly transform the way companies approach resource management, maintenance and process optimization.

Process modelling with fuzzy logic (integrating supply, production and deadline uncertainties) analysed and processed by an adapted AI is now a formidable method for designing “Design for Lean”.? It is also possible to continue capturing information, integrating their real-time variations in a module and thus achieving continuous process improvement.

Sources:

• Zadeh, L. A. (1965). ?Fuzzy sets. Information and Control, 8(3), 338-353.
• Mohagheghi, S., & Front, A. (2017). Cyber-physical systems for smart factories. International Journal of Computer Integrated Manufacturing, 30(4-5), 396-411.
• Negi, D.S. & Jain, A. (2020). ?A Survey of Fuzzy Logic Applications in Industry 4.0. Journal of Intelligent & Fuzzy Systems, 39(6), 8343-8354.