Programmable logic controllers (PLCs) have become an indispensable part of automated production in professional practice. Their programming is, among other things, an integral part of learning factories in the context of Industry 4.0 (Wilbers & Windelband, 2021) and therefore essential for apprentices in the field of automation technology/mechatronics in Europe.

Approaches to programming, for example in computer science didactics, are often based on the theory of problem-solving (Adelson, 1981; Dewey, 2002; Dörner, 1976). Thus, an initial problem to be solved can be regarded as the initial state of a program through a suitable representation (Anderson, 2013). The solution process is carried out by performing operations that are defined in the program with the aim of achieving a defined end state. In order to achieve this final state, suitable operations must be found for the solution process, whereby corresponding specialised technical concepts from programming must be used as problem-solving operators (Wieczorek, Ribe, Class & Brinkmeier, 2017).

Previous studies showed a clear discrepancy between the performance actually achieved and the level required in the curriculum with regard to programming a PLC. After three years of training, 64% of trainees in the occupation electronic technician for automation technology and mechatronic technician were not able to independently solve PLC programming problems that are taught in the middle of the second year of training (Author1, 2016). The identification of such obstacles to understanding and the awareness of their characteristics in the didactic and pedagogical process offers the possibility for future teaching practice "to recognize problems of understanding, to weigh possible reactions, to test them and to reconsider them" (Schecker & Duit, 2018, p. 7; translated by the authors). Our main research aim is to examine in greater detail what concrete obstacles occur for trainees when programming a PLC.

The research methodology for the systematic analysis of concrete obstacles in PLC programming is based on an empirical-qualitative design of Author2 (2011). In the context of this article, a baseline scenario is presented, which was developed on the basis of curricular analyses as well as in consultation with experts of vocational training practices (teachers, training staff, representatives of an audit institution). Thus, problems with different focuses (reversing contactor circuit, sequencer, analog value processing) must be programmed with different operators (such as AND, OR, SR/RS-Flipflop). Based on qualitative content-analytical evaluation (Gläser & Laudel, 2009; Mayring, 2019), an inductive category formation is carried out by analysing the programming tasks. For the purpose of data collection, problem-solving processes for a total of n = 150 electronic technicians for automation technology and mechatronic technicians in Germany were examined at the end of the vocational training. The categories are derived directly from the existing survey material in this context written student responses (programming code). Mayring (2019) calls this type of category formation a summarizing content analysis with the goal of narrowing down the text elements without distorting the core content of the material. To calculate intercoder reliability for nominal scaled data, 10% of the data are double coded by a second independent person. In the context of the paper, the widely used liberal overlap measure of the Holsti coefficient is computed (Döring & Bortz, 2016; Kolb, 2014). For the coded data, a satisfactory Holsti coefficient of rH = .96 can be reported.

The analyses so far show that it is possible to identify and characterize concrete obstacles in PLC programming in selected problem areas. For example, the problem "reversing contactors" (n = 112) indicates that the TOP3 obstacles of learning are connected to the imprecise designation of program components e.g. omitting Quit for SR FlipFlop (n = 75). The negation of the end switch B9 and or B10 was incorrectly executed (n = 69) and the lock key for clockwise/counterclockwise rotation is not properly functional (n = 52). This is a security problem as contactors could simultaneously toggle. The problem "quantity control" (n = 124) shows that the majority of trainees designate the program blocks (n = 56) unclearly, implement incorrectly the emergency stop or the confirm variable (n = 50) and use the wrong input for the compare component (n = 43). In particular, when it comes to information technology content in the context of PLC programming, understanding obstacles arise according to the analyses to date. This is at variance with the expected strong influence of information technology on Industry 4.0 production. The exact analysis of the programming tasks will be completed in autumn. The obstacles to understanding identified and characterized here should now be used to develop best-practice teaching models for programming a PLC.

Adelson, B. (1981). Problem solving and the development of abstract categories in programming languages. Memory & Cognition, 9(4), 422–433. https://doi.org/10.3758/BF03197568 Anderson, J. R. (2013). Kognitive Psychologie (Lehrbuch, 7., erw. und überarb., neu gestaltete Aufl.). Berlin: Vs Verlag Fur Sozialwissenschaften. Dewey, J. (2002). Wie wir denken (John-Dewey-Reihe, Bd. 2). Zürich: Pestalozziamu. Dörner, D. (1976). Problemlösen als Informationsverarbeitung (Kohlhammer Standards Psychologie: Studientext : Teilgebiet Denkpsychologie). Stuttgart, Berlin, Köln, Mainz: Kohlhammer. Gläser, J. & Laudel, G. (2009). Experteninterviews und qualitative Inhaltsanalyse als Instrument rekonstruierender Untersuchungen. Wiesbaden: VS. Mayring, P. (2019). Qualitative Inhaltsanalyse. Grundlagen und Techniken. Wieselburg: Julius Beltz. Schecker, H. & Duit, R. (2018). Schülervorstellungen und Physiklernen. In H. Schecker, T. Wilhelm, M. Hopf & R. Duit (Hrsg.), Schülervorstellungen und Physikunterricht (S. 1-21). Berlin: Springer Spektrum. Wieczorek, B., Ribe, L., Class, C. B. & Brinkmeier, M. (2017). Analogien für Programmierkonzepte: Ein Weg zum Computational Thinking. In I. Diethelm (Hrsg.), Informatische Bildung zum Verstehen und Gestalten der digitalen Welt. 17. GI-Fachtagung Informatik und Schule (S. 349–358). Bonn. Wilbers, K. & Windelband, L. (Hrsg.). (2021). Lernfabriken an beruflichen Schulen. Gewerblich-technische und kaufmännische Perspektiven (Blaue Reihe, Bd. 26, 4. Auflage). Berlin: epubli.