Arcelik specializes in consumer durables and consumer electronics, offering production, marketing, and after-sales support services in these sectors. Arcelik is Europe’s fourth-biggest white goods company in total sales.
Having operations in durable consumer goods industry with production, marketing and after-sales services, Arçelik A.Ş. offers products and services around the world with its 30,000 employees, 23 different production facilities in 9 countries, its 35 sales and marketing companies in 33 countries all over the world and its 12 brands serving products and services in more than 146 countries. The annual production volume of Arçelik is more than 20 million units.
Arçelik attributes its success to its significant investments in R&D and technology. Providing customers with technologically innovative products is a key aspect of Arçelik's strategy to strengthen its brands and its global presence. Arçelik has leveraged product specific R&D across its segments in order to further drive innovation. Arçelik seeks to ensure that its intellectual property is protected with patent applications. Arçelik is IPR leader in Turkey and ranked at top 100 in WIPO’s Top Patent Applicants list.
The electric motor of white good appliances is manufactured by Arçelik itself. In functional test system of produced motors is based on automatic calculated special power spectrum energies, is compared with threshold values by determined Vibration R&D Lab and pass/fail decision is finalized according this comparison.
However, there are cases that even the decision is fail, some of the failures cannot be detected by the system and as a reason to that, final decision is dependent to the operator. Operator need to listen the noise of the motor and make the final decision of pass/ fail. The goal is to create a new test vehicle performing fully automatic quality test of electric motor and develop AI based feedback algorithm from aging test for correction and prediction of optimized threshold values.
Implementation and Designing Data Collection Systems and Infrastructures & Improve reliability checkers & Deployment AI based algorithms (anomaly detection, fault detection, remaining useful life estimation):
With creating edge-to-edge connection between production test vehicle and aging test vehicle, we want to create AI based test system to improve our test vehicles reliability that will help us to understand how a motor driver should be designed to provide diagnostic and prognostic information on electric motor. To collect signals with high resolution, a data acquisition platform will be installed.
Besides vibration data, current, voltage and sound data will be collected in this new platform for analysis. Collected data will be feed to a deep learning algorithm to optimize classification in production test vehicle.
Online deep learning algorithm and similar data acquisition platform will be applied to aging test vehicle to give feedback for threshold of production test vehicle.
The most important features calculated from raw data are founded by using feature selection methods (like max relevance min redundancy MRMR) that will give an input to new motor driver design.
The most essential features of electric motor will be learnt with help of AI based test system.
To be able to provide these features, some electronic components on motor driver will be needed. Hence a new motor driver that gives information on diagnostic and prognostic of electric motors will be able to design.