Knowledge-based system to support product development focusing on diagnosis of low performance in hermetic compressors

• Published in: Journal of the Brazilian Society of Mechanical Sciences and Engineering 2015-11, Vol.37 (6), p.1731-1742
• Main Authors: Pedroso, Altemir Palmede, da Silva, Jonny Carlos
• Format: Article
• Language: English
• Subjects: Engineering; Mechanical Engineering; Technical Paper
• ISSN: 1678-5878; 1806-3691
• DOI: 10.1007/s40430-014-0279-z

In an industrial context, performance testing analysis is conducted during the final product development phase, such analysis has been carried out by product analysts. Based on a multinational compressor manufacturer expertise, this activity was considered a candidate for a knowledge engineering initiative. This paper presents the development of a knowledge-based system prototype for diagnosis of low performance root causes in hermetic compressors. Diagnosis techniques were implemented to investigate the main compressor performance attributes: cooling capacity, power consumption, minimum starting voltage and noise. This work documents the Knowledge Engineering process. Conclusions demonstrate that this methodology was applied with relevant results for root cause diagnosis of hermetic compressors, which contributes to improve the product development cycle. Moreover, the diagnoses generated by the prototype are updated and aligned with the recommendations of experts from Brazil and China, who validated the prototype results. The prototype contributes to improve fault diagnosis quality and reduces analysis time and cost. Two main outcomes were that the system has been incorporated as part of TechSkill program to train new products analysts and experts from different countries have requested to use system.