On December 4, 2020, the first MathTech Workshop on Mathematical Technologies for Industry was held in Poland. It was organized by the PL-MATH-IN network as a satellite event of the annual EU-MATH-IN Council Meeting. Originally, the workshop was planned to be held at the Institute of Mathematics of the Polish Academy of Sciences (IMPAN) in Warsaw. However, due to the severe epidemic situation in Poland, it took place online.
The goal of the workshop was to report on recent interesting applied maths projects carried out in Poland. The first edition had five speakers. The first speaker was Agnieszka Wyłomańska from Wrocław University of Science and Technology (WUST). She presented new mathematical methods for damage detection in mining machines.
Professor Wyłomańska has a long experience in cooperating with the KGHM mining company in the area of local damage detection of mining machines. Both the Faculty of Pure and Applied Mathematics and Faculty of Geoengineering, Mining, and Geology (WUST) are involved in the EiT Raw Material Project OPMO “Operational Monitoring of Mineral Crushing Machinery”, where new advanced mathematical techniques are designed for monitoring systems of mining machines. Those methods are based on the vibration signal analysis where additional disturbances (not related to damage) may occur. The perfect example is a crushing machine that produces disturbances of large amplitudes much higher than the background noise. In such a case the local damage identification (that is manifested by cyclic impulses hidden in the noise) is a very difficult task. The classical methods based on the measures of impulsiveness fail in this case. Recently, a new approach proposed by the research group from WUST has been applied. The new methodology assumes that the vibration signal is described by a heavy-tailed class of distributions. This approach seems to be promising in the considered problem and gives much better results than the classical tools used in the literature. The presented results have been already published in high-impact journals devoted to signal processing methods used in mechanical systems.
The next speaker was Witold Pawlus from Nokia AI Lab who introduced a new forecasting method of raw materials deliveries in supply chain management. Paweł Dłotko from IMPAN gave an overview of problems in applied science that can be solved using methods from computational topology and, in general, computational mathematics. Piotr Miłoś from IMPAN reported on recent advances in reinforcement learning methods for autonomous driving systems. Last but not least, Marek Niezgódka from Cardinal Wyszyński University in Warsaw presented a summary of his maths-driven industrial projects. All talks were accompanied by follow-up questions and discussion to retain some characteristics of a traditional workshop.
The workshop was openly accessible to the public. Although the online-only form of such a meeting may be still perceived as a little bit unusual, it was a success, attracting many researchers, engineers, and entrepreneurs from across Europe. Since all attendees avoided air travel and cost, CO2 emissions and also reduced the impact on families, such a form may turn out to be particularly beneficial in the future, not only in the phases of public health emergencies.
Polish Mathematical Society
Chair of PL-MATH-IN