• Artificial intelligence is successfully conquering the conductor plate industry
• Automated image recognition and analysis are stepping stones for cost reduction in terms of quality control and resources.
• Explainable AI takes aim at strengthening trust and acceptance of smart systems

Conductor plates are known as the nervous system of any electronic device. No matter if it is about mobile end devices or the automotive, industrial and medical sectors. The areas of application are diverse, as is their production. A single conductor plate requires around 150 complex steps until it is ready for use.

Quality control made possible at highest level

“At AT&S high quality is a given when it comes to our products. Conductor plate images are taken automatically during the process of manufacturing and subsequently pass through image analysis software. Sometimes conductor plates are falsely identified “defective”. Regrettably for us, without any comprehensible reasons. This may result in additional loss of time and resources.” Ulrike Klein states, Head of Data & Analytics at AT&S.

Know Center developed an AI-algorithm for AT&S, its project partner and leading manufacturer of high-end conductor plates. This algorithm not only identifies images of conductor plates properly, but in addition provides an explanation as to why a conductor plate has been identified defective or intact. AT&S consequently owns a transparent AI system, which will deliver comprehensible and explainable results in the foreseeable future once intensively tested.

Technicians and AI consent

“Our goal was to precisely detect faulty PCBs and make the results traceable. We are pleased we succeeded by implementing the project and our results also agreed with the propositions made by AT&S’s technicians,” Dr. Andreas Trügler states, Head of Research, module DDAI, at Know-Center, and further explains: “First, our algorithm had to understand which PCBs were faulty and why. In order to achieve this, a neural network was trained and supplied with image data of correct and faulty conductor plates by our team. By using methods from the Explainable AI research field, we were able to provide additional explanation in terms of why and where a PCB was identified as faulty.”

Smart production environment using AI

Industry 4.0 or so-called “Smart Factory” no longer is a dream. The insert of smart machines and applications provides companies with significant advantages in times of increasingly competitive pressure.

Now more than ever AI is the driving technology for innovative products and services in the digital age. This fact primarily has become noticeable for the manufacturing industry in the field of automation. Stefanie Lindstaedt, Managing Director at Know-Center explains: “AI enables quality assurance at the highest level and thereby helps organisations save costs and resources. However, quality gaps still exist when it comes to automated image recognition and analysis, which keeps advancing on various sectors of the industry. Building trust in these technologies proved to be another obstacle on the path to firmly embedding AI within organizations. We are very pleased this project has succeeded by overcoming both.”

AT&S Austria Technologie & Systemtechnik AG, is one of Know Center’s industrial partners joint by the module DDAI which runs under the auspices of COMET. The module, which is led by Know-Center and funded by FFG, aims to develope safe, verifiable and explainable AI which assures privacy protection simultaneously. It will significantly contribute to acceptance and trust in AI. In the future more projects leading to “trustworthy AI” will be promoted collectively by AT&S and other industry partners which are part of the module.

About Know-Center

Know-Center is one of the leading European research centers for Data-driven business and AI. It has been supporting well-known companies by using data as a factor of success for their businesses since 2001. Know-Center relies on established Big Data as well as High Performance Computing (HPC) infrastructure for data analysis. Known as constant part of the European research landscape, the center very successfully processed numerous projects and contract research on an European as well as domestic level. The K1 Competence Center, which is funded within the framework of COMET, is the leading training center for data scientists in Austria and in addition offers a range of Al training and consulting services to companies. The majority shareholder is Graz University of Technology, a key supporter of domestic AI research, whose institutes process numerous projects together with Know-Center. In 2020, Know-Center was the only Austrian center to receive the EU’s iSpace Gold Award, which only has been awarded nine times in the entire EU so far.
Further Information: www.know-center.at

About AT&S

AT&S is one of the world’s leading manufacturers of high-quality conductor plates and IC substrates. AT&S industrializes trendsetting technologies for its core businesses mobile devices, automotive, industry, medicine and advanced packaging. AT&S, an international high-growth company commands a global presence, with production sites in Austria (Leoben, Fehring) and plants in India (Nanjangud), China (Shanghai, Chongqing) and Korea (Ansan near Seoul).

• New study determined the accuracy of algorithms based on four common music recommendations
• Mainstream music listeners appeared to receive more accurate music recommendations than non-mainstream music listeners
• Findings could improve music recommendation systems

According to a research study published in the renowned open access journal EPJ Data Science, music recommendations for fans of non-mainstream music, such as hard rock and ambient, may be less accurate than for those of mainstream music, such as pop.

A team of researchers at Know-Center, Graz University of Technology, Johannes Kepler University Linz, University of Innsbruck and University of Utrecht (the Netherlands) analysed the accuracy of algorithm-generated music recommendations for listeners of mainstream and non-mainstream music. They used a dataset containing the listening histories of 4,148 users of the music streaming platform Last.fm who listened mostly to non-mainstream music or mostly mainstream music.

„Since more and more music is becoming available via music streaming services, music recommendation systems have become essential for helping users to search, sort and filter extensive music collections. However, the quality of many cutting-edge music recommendation techniques for non-mainstream music users still leaves a lot to be desired. In our study we established that the willingness of a music user to listen to the kinds of music outside of his or her primary music preferences has a positive effect on the quality of recommendations. Thinking out of the box pays off even when it comes to listening to music,“ explained Dominik Kowald, first author of the study and Head of Research Area Social Computing at Know-Center.

Algorithm categorizes music listeners

“First, we established a computational model for artists, to whom music users listened most frequently. Based on this model, we were able to predict the probability of music users’ liking the music recommended to them by four common music recommendation algorithms. We found that listeners of mainstream music appeared to receive more accurate music recommendations than listeners of non-mainstream music“, stated Elisabeth Lex, the study’s scientific leader and Associate Professor of Applied Computer Science at Graz University of Technology.

Non-mainstream music listeners have been assigned by an algorithm into the following categories based on the types of music they most frequently listened to: users of music genres with only acoustic instruments (e.g. folk), users of high-energy music (e.g. hard rock, hip-hop), users of music with acoustic instruments and no vocals (e.g. ambient), and users of high-energy music with no vocals (electronica).

By comparing each group’s listening histories, the researchers identified users who were most likely to listen to music outside of their preferred genres and the diversity of music genres listened to by each group. Those who mainly listened to music such as ambient were found to be most likely to listen to music preferred by hard rock, folk and electronica listeners as well. Those who mainly listened to high-energy music were least likely to listen to music preferred by folk, electronica and ambient listeners. However, they listened to the widest variety of genres, e.g. hard rock, punk, singer/songwriter and hip-hop.

Biased music recommendation algorithms

By using the computational model, the researchers predicted the probability of various groups of non-mainstream music listeners’ liking the music recommendations generated by the four common music recommendation algorithms. They found that those who listened mainly to high-energy music appeared to receive the least accurate music recommendations and those who mainly listened to ambient get the most.

Stefanie Lindstaedt, CEO of Know-Center and Director Institute of Interactive Systems & Data Science (ISDS) at TU Graz stated: „At Know-Center we research smart recommender algorithms for various contents and domains. We often face the problem of insufficient recommendations generated for users who have no mainstream preferences or for rarely used contents. However, we have made significant progress and will integrate the study’s findings into our recommender offers. Moreover, we would like to put this knowledge to good use by reducing discrimination potentials of algorithms in general in order to advance the development of trustworthy artificial intelligence within Austria and Europe.“

The authors pointed out potential benefits of their findings for the development of music recommendation systems that will provide more accurate recommendations to non-mainstream music listeners. However, they caution that since the study was based on a sample from Last.fm users, the results may not be representative of all Last.fm users and users of other music streaming platforms.

Study: Dominik KOWALD, Peter MUELLNER, Eva ZANGERLE, Christine BAUER, Markus SCHEDL, Elisabeth LEX. Support the Underground: Characteristics of Beyond-Mainstream Music Listeners. EPJ Data Science (2021)
https://doi.org/10.1140/epjds/s13688-021-00268-9

Contact:

Kowald, Dominik, Dipl.-Ing. Dr.techn. BSc
Know-Center | Research Area Manager Social Computing @ Know-Center
Phone: +43 316 873 30846
e-mail:

Elisabeth Lex, Assoc.Prof. Dipl.-Ing. Dr.techn.
TU Graz | Institut für Interaktive Systeme und Data Science
Phone: +43 316 873 30841
e-mail: