The Application of neural networks to modular arrangements of predetermined time standards

Basitere, Emmanuel (avtor), Daniyan, Ilesanmi (avtor), Mpofu, Khumbulani (avtor), Adeodu, Adefemi (avtor)

Modular arrangements of predetermined time standards (MODAPTS) is an effective and efficient method to measure work and the activities associated with it. The time standard is used all over the world in different industries, but the method is old, slow, and difficult for first-time users to work with. This study applied neural networks (NN) to MODAPTS. Primary training data in the form of MODAPTS keywords were employed. The training data were acquired as raw data in the form of MODAPTS time studies. These data however was then broken and processed to extract the keywords for the training of the NN. The NN was also trained with the data collected using the TensorFlow algorithm assisted by the Keras library. This was achieved by first learning the fundamentals of creating a NN. Thereafter, consolidating several tools, such as the Python programming language and the Keras library, were used to implement the artificial neural network (ANN). The results obtained indicated that 94.7 % of successful predictions were made by the NN while only 5.3 % were manually entered codes to correct the ANN chatbot. The mean difference between the two methods is 0.25 minutes; the t-test was calculated at 95 % confidence level (0.05) and produced a P-value of 0.9663. The computed P-value was greater than 0.05, showing that there is no significant difference between the two generated studies. The MODAPTS-ANN technique demonstrated in this study possesses great potential to improve and refine work measurement.

• Jezik: Angleški jezik
• Ključne besede: artificial neural network, modular arrangement, predetermined time standards, TensorFlow algorithm
• Vrsta gradiva: Članek v reviji
• Tipologija: 1.01 - Izvirni znanstveni članek
• Organizacija: FS - Fakulteta za strojništvo
• Status publikacije: Objavljeno
• Različica publikacije: Objavljena publikacija
• Leto izida: 2023
• Št. strani: Str. 61-72
• Številčenje: Vol. 69, no. 1/2
• PID: 20.500.12556/RUL-144951
• UDK: 681.5:004.021
• ISSN pri članku: 0039-2480
• DOI: 10.5545/sv-jme.2022.238
• COBISS.SI-ID: 146543619
• Datum objave v RUL: 24.03.2023

Gradivo je del revije

• Naslov: Strojniški vestnik
• Skrajšan naslov: Stroj. vestn.
• Založnik: Zveza strojnih inženirjev in tehnikov Slovenije [etc.], = Association of Mechanical Engineers and Technicians of Slovenia [etc.
• ISSN: 0039-2480
• COBISS.SI-ID: 762116

Licence

• Licenca: CC BY 4.0, Creative Commons Priznanje avtorstva 4.0 Mednarodna
• Povezava: http://creativecommons.org/licenses/by/4.0/deed.sl
• Opis: To je standardna licenca Creative Commons, ki daje uporabnikom največ možnosti za nadaljnjo uporabo dela, pri čemer morajo navesti avtorja.

Sekundarni jezik

• Jezik: Slovenski jezik
• Naslov: Uporaba nevronske mreže pri modularni ureditvi vnaprej določenih časovnih normativov
• Ključne besede: umetna nevronska mreža, modularna ureditev, vnaprej določeni časovni normativi, algoritem TensorFlow