Modelling Production Uncertainties Using The Adaptive Neuro-Fuzzy Inference System

doi:10.7166/26-1-560

Study

Commercial ProductionHigh ImpactStrong effect

ANFIS models production throughput with 98% accuracy under uncertainty

An adaptive neuro-fuzzy inference system (ANFIS) can accurately predict production throughput by accounting for common manufacturing uncertainties.

The South African Journal of Industrial Engineering · 2015

01

Key Findings

01The ANFIS model demonstrated a high capability in forecasting production throughput under uncertainty.
02The ANFIS model achieved an R-squared value of 98%, significantly outperforming the MLR model.

02

Application

Design takeaway

Implement ANFIS or similar intelligent systems to model and predict production throughput, incorporating key uncertainties for more robust operational planning.

How to apply

Use ANFIS to build predictive models for production lines, feeding in data on scrap rates, machine downtime, demand forecasts, and material lead times to estimate future output.

Project actions

01When modelling production, identify all potential sources of uncertainty.
02Explore advanced computational techniques like ANFIS for more accurate predictions.

03

Method & Evidence

AimCan an adaptive neuro-fuzzy inference system (ANFIS) effectively model production throughput under various uncertainties with higher accuracy than traditional statistical methods?

MethodComparative modelling study

ProcedureAn ANFIS model was developed to predict the production throughput of floor tiles, considering five key uncertainties: scrap, setup time, break time, demand, and lead time. The model's performance was evaluated using 104 weekly observations and compared against a multiple linear regression (MLR) model.

Sample104 observations

ContextManufacturing industry (floor tile production)

Variables

IV["Scrap rate","Setup time","Break time","Demand","Lead time"]

DVProduction throughput

CV["Production system (floor tile manufacturing)","Observation period (weekly)"]

04

Strengths & Limitations

Strengths

+High predictive accuracy achieved by the ANFIS model.
+Direct comparison with a standard statistical method (MLR) provides a clear benchmark.

Limitations

The ANFIS model's performance is dependent on the quality and completeness of the input data. It may also require significant computational resources for training.

Reliability & validity

The study's reliability is supported by the use of a defined dataset and a comparative analysis. Validity is enhanced by the high R-squared value, indicating the model effectively explains the variance in production throughput.

Think critically

How might the 'black box' nature of ANFIS affect the interpretability of its predictions in a real-world production setting, and what strategies could be employed to address this?

05

Design Principles

"Predictive modelling of complex systems should account for inherent uncertainties to improve accuracy and decision-making."

Understanding and predicting production throughput is crucial for efficient manufacturing operations. By incorporating uncertainties like scrap, setup times, and demand fluctuations into predictive models, businesses can better manage resources, optimize schedules, and mitigate potential disruptions.

06

What This Means for Your Design

This study shows that a smart computer system called ANFIS can predict how much a factory will produce much better than older methods, even when things like machine breakdowns or changing customer orders make it hard to know for sure.

How to use in your project

1.Use the findings to justify the selection of advanced modelling techniques for predicting project outcomes.
2.Reference the study when discussing the challenges of uncertainty in production planning.

07

Add to My Project

08

Quick Cite

(2015). MODELLING PRODUCTION UNCERTAINTIES USING THE ADAPTIVE NEURO-FUZZY INFERENCE SYSTEM. The South African Journal of Industrial Engineering. https://doi.org/10.7166/26-1-560 Retrieved from https://designdex.org/study/c36da9c8-ce41-42b2-981e-d5cdfd993a00/anfis-models-production-throughput-with-98-accuracy-under-uncertainty

Paragraph starter

This research demonstrates the efficacy of the Adaptive Neuro-Fuzzy Inference System (ANFIS) in modelling production throughput under uncertainty, achieving a remarkable 98% accuracy. This highlights the potential for advanced computational intelligence to significantly improve the predictability of manufacturing processes, a crucial factor in efficient resource allocation and operational planning.

09

Source

The South African Journal of Industrial Engineering

MODELLING PRODUCTION UNCERTAINTIES USING THE ADAPTIVE NEURO-FUZZY INFERENCE SYSTEM

journal · 2015

View source

Questions about this research

What does the research say about anfis models production throughput with 98% accuracy under uncertainty?: Implement ANFIS or similar intelligent systems to model and predict production throughput, incorporating key uncertainties for more robust operational planning. Evidence: The South African Journal of Industrial Engineering (2015).
Why does "ANFIS models production throughput with 98% accuracy under uncertainty" matter for design?: Understanding and predicting production throughput is crucial for efficient manufacturing operations. By incorporating uncertainties like scrap, setup times, and demand fluctuations into predictive models, businesses can better manage resources, optimize schedules, and mitigate potential disruptions.
How can designers apply this research?: Implement ANFIS or similar intelligent systems to model and predict production throughput, incorporating key uncertainties for more robust operational planning.
What were the main findings?: The ANFIS model demonstrated a high capability in forecasting production throughput under uncertainty.. The ANFIS model achieved an R-squared value of 98%, significantly outperforming the MLR model.
What research method was used?: Comparative modelling study with 104 observations.
How strong is the evidence?: Evidence strength is rated Strong effect, based on a 2015 journal from The South African Journal of Industrial Engineering.
What should I do differently in my next project?: Use ANFIS to build predictive models for production lines, feeding in data on scrap rates, machine downtime, demand forecasts, and material lead times to estimate future output.
What are the limitations?: The study focused on a specific product (floor tiles) and a defined set of uncertainties; generalizability to other industries or a broader range of uncertainties may vary.
Is there evidence that production throughput affects design outcomes?: The ANFIS model predicted production output with 98% accuracy, showing it's a superior tool for forecasting in uncertain manufacturing environments compared to traditional regression. Understanding and predicting production throughput is crucial for efficient manufacturing operations. By incorporating uncertainties lik Source: The South African Journal of Industrial Engineering (2015).
Where does this anfis research apply?: Manufacturing industry (floor tile production) It sits within commercial production research on designdex.org.