Lead time prediction using machine learning algorithms: A case study by a semiconductor manufacturer

Lingitz, Lukas; Gallina, V.; Ansari, Fazel; Gyulai, Dávid; Pfeiffer, András; Sihn, Wilfried; Monostori, László

doi:10.1016/j.procir.2018.03.148

Cited by 105 publications

(58 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The set of input features in this case could be similar to the ones aforementioned (process related) except that the target label would be binary (0 in case of normal operation and 1 in the case of machine breakdown or failure). Depending on the type of algorithm, i.e., either regression or classification, different types of metrics exist to gauge the accuracy of the model [35,36] and thereby improve the performance of the ML model either during cross validation or during subsequent training postiterative feature engineering.…”

Section: Basics Of ML In Manufacturingmentioning

confidence: 99%

“…ML algorithms have the ability to learn from historical or real-time data to predict events, which can help in informed decision-making in the domain of production and supply chain planning, scheduling, and control. A detailed review on ML applications to production planning and control is presented in a recent article by Cadavid et al [39] Lingitz et al [35] used the historical data from the MES for a period of 2 years and conducted a comparative evaluation of 11 supervised ML algorithms to predict the lead time in a semiconductor production facility. Among the various algorithms evaluated, including LR, RR, lasso regression, multivariate adaptive regression (MARS), regression tree (RT), bagged RT, RF, boosted RF, SVM, k-NN, and ANN, it was observed that the RF algorithm gave the best prediction whereas ANN had the least predictive ability.…”

Section: For Production and Supply Chain Planningmentioning

confidence: 99%

See 1 more Smart Citation

Smart Manufacturing for Smart Cities—Overview, Insights, and Future Directions

Suvarna

Büth

Hejny

et al. 2020

Advanced Intelligent Systems

View full text Add to dashboard Cite

According to statistics provided by the UN in 2018, the global urban population rose from 0.75 billion in 1950 to 4.2 billion by 2018, a 5.5-fold increase in nearly seven decades. Currently, this 4.2 billion urban population accounts for 55% of the current global population, and given the trend, the global urban population is expected to touch 68% of total population by 2050. [1,2] To keep abreast with this growing shift toward urbanization, the last decade witnessed a multitude of global projects across the USA, EU, Asia Pacific, and Middle East regions to cater to the needs of a futuristic urban population through the concept of "smart cities." [3,4] Some of the popular programs include Smart Cities Initiatives by the USA, [5] EU H2020 Smart Cities and Communities project, [6] Smart Nation Program by Singapore, [7] Global Initiative for Resource Efficient Cities (GI-REC) by UN, [8] etc., to mention a few. Existing literature shows that there is no single or standardized definition to the concept of a smart city. [9] However, the central theme usually revolves around the application of information and digital technologies to provide intelligent and innovative solutions to the demanding requirements of an urban ecosystem including infrastructure, transport, healthcare, governance, and security. [2,10] The overall essence and motivation of a smart city is to facilitate the highest quality of life to its residents, while simultaneously optimizing the resources and energy efficiency, aiding the overall sustainability, social, and economic development of the city. Given the fact that the concept of smart cities is constantly evolving, devising a single template to describe the various characteristic features of a smart city can be a challenging task. [9] However, notable efforts have been made to determine and define a few salient features or indices of a smart city, published by research organizations as well as by consulting firms. [11] Both types of reports present commonalities in terms of indices, which are integral to smart cities and hence enable their acceptance with a degree of confidence. Some of the most notable and cited reports include Cities in Motion Index by IESE, [12] Smart Cities Ranking by Juniper Research, [13] Top 50 Smart City

show abstract

Section: Basics Of ML In Manufacturingmentioning

confidence: 99%

Section: For Production and Supply Chain Planningmentioning

confidence: 99%

Smart Manufacturing for Smart Cities—Overview, Insights, and Future Directions

Suvarna

Büth

Hejny

et al. 2020

Advanced Intelligent Systems

View full text Add to dashboard Cite

show abstract

“…Machine learning methods have been used before for analysis and decision making in semiconductor fabs. Some examples are work-in-progress prediction [13], lead time prediction [14], dynamic storage dispatching [15], vehicle traffic control [16], and wafer defect detection using image classification [17,18]. The machine learning method is classified as a data-driven approach that is suitable for cases with complicated relationships between many factors [19].…”

Section: Literature Reviewmentioning

confidence: 99%

Production Flow Analysis in a Semiconductor Fab Using Machine Learning Techniques

Singgih

2021

Processes

View full text Add to dashboard Cite

In a semiconductor fab, wafer lots are processed in complex sequences with re-entrants and parallel machines. It is necessary to ensure smooth wafer lot flows by detecting potential disturbances in a real-time fashion to satisfy the wafer lots' demands. This study aims to identify production factors that significantly affect the system’s throughput level and find the best prediction model. The contributions of this study are as follows: (1) this is the first study that applies machine learning techniques to identify important real-time factors that influence throughput in a semiconductor fab; (2) this study develops a test bed in the Anylogic software environment, based on the Intel minifab layout; and (3) this study proposes a data collection scheme for the production control mechanism. As a result, four models (adaptive boosting, gradient boosting, random forest, decision tree) with the best accuracies are selected, and a scheme to reduce the input data types considered in the models is also proposed. After the reduction, the accuracy of each selected model was more than 97.82%. It was found that data related to the machines’ total idle times, processing steps, and machine E have notable influences on the throughput prediction.

show abstract

“…[3] developed a tree-based piecewise linear regression model to estimate the flow-time of a manufacturing system. [4] used different machine learning approaches to improve the lead time prediction for a mixed dataset from a manufacturing execution system. Tree-based ensemble methods have the lowest root mean square error (RMSE) and mean absolute error (MAE).…”

Section: Literature Reviewmentioning

confidence: 99%

Using Partial Combination Models to Improve Prediction Quality and Transparency in Mixed Datasets

Chang

Tien

et al. 2020

IEEE Access

View full text Add to dashboard Cite

Mixed Datasets with complex interactions between categorical and numerical attributes are common in engineering and business applications. For example, production rates in manufacturing systems are jointly influenced by several categorical and numerical attributes, such as machine and product types and their numerical attributes. This study aims to improve the prediction performance and transparency of mixed datasets with complex interactions using machine learning (ML) methods. The proposed method requires lesser data and computational effort than existing hierarchical or clustering regression methods. Multiple prediction models can be generated by partitioning a dataset into subsets with different categorical attribution combinations. One-and two-stage model selection methods are proposed to use the training and validation datasets in selecting better models among all the prediction models. Numerical results demonstrate the potential of the model selection approach in a mixed dataset from a semiconductor manufacturer. In comparison with regression models, more than 30% reduction in root mean square error is observed using the proposed model selection approach. The cross-validation test results also demonstrated a 10% improvement in accuracy against the properly tuned XGBoost models. Moreover, the proposed model selection approach is compatible with other regression or ML prediction methods and can be used to improve the model's transparency of any existing methods on mixed datasets.

show abstract

Lead time prediction using machine learning algorithms: A case study by a semiconductor manufacturer

Cited by 105 publications

References 15 publications

Smart Manufacturing for Smart Cities—Overview, Insights, and Future Directions

Smart Manufacturing for Smart Cities—Overview, Insights, and Future Directions

Production Flow Analysis in a Semiconductor Fab Using Machine Learning Techniques

Using Partial Combination Models to Improve Prediction Quality and Transparency in Mixed Datasets

Contact Info

Product

Resources

About