Hybrydowy model priorytetyzacji technologii

W monografii zaprezentowano hybrydowy model priorytetyzacji technologii będący oryginalną propozycją rozwiązania złożonego zagadnienia zorientowanej na przyszłość porównawczej oceny technologii przeprowadzanej w sytuacji braku możliwości jednoznacznego określenia parametrów i potencjału technologii oraz wynikającej z tego niepewności. W modelu zaproponowano synergiczne połączenie metod zbiorów przybliżonych (Rough Set) i analizy obwiedni danych (Data Envelopment Analysis – DEA). Prezentację struktury modelu poprzedza przegląd i synteza stanu wiedzy w zakresie metod dotychczas stosowanych w ocenie technologii. Zidentyfikowany brak badań i wiedzy dotyczących wykorzystania modeli hybrydowych w procesie priorytetyzacji technologii – a w szczególności łączących teorię zbiorów przybliżonych oraz metody badań operacyjnych w celu zwiększenia obiektywizmu oceny oraz ograniczenia arbitralności decyzyjnej przy ujmowaniu subiektywnych opinii – był motywem opracowania i zbadania możliwości hybrydowego modelu. W monografii przedstawiono także fundamentalia teorii zbiorów przybliżonych i metody DEA tworzące teoretyczną podbudowę ich syntezy w hybrydowym modelu.

The monograph presents a hybrid model of technology prioritisation, which is an original proposal to solve the complex issue of a comparative future-oriented technology assessment carried out in the situation of vague determination of parameters, technology potential and the resulting uncertainty. The model proposes a synergistic combination of Rough Sets and Data Envelopment Analysis (DEA). The presentation of the model structure is preceded by a review and synthesis of state of the art in the field of methods that have been used in technology assessment. The identified lack of research and knowledge regarding the use of hybrid models in the process of technology prioritisation – and, in particular, integrating Rough Sets theory and operational research methods to increase the objectivity of evaluation and reduce decision-making arbitrariness while analysing subjective opinions, was the motivation to develop and explore the capabilities of the hybrid model. The monograph also presents the foundations of the Rough Set Theory and DEA methods that form the theoretical framework for their synthesis in the hybrid model. The hybrid model uses: (i) the concept of reducts from the Rough Set Theory to limit the number of criteria; (ii) lower and upper approximation of the sets to model uncertainty; (iii) linear optimisation algorithms from the DEA method for objective determination of criteria weights; (iv) the relation of the DEA method efficiency to formulate the assessment score. The proposed assessment process by the hybrid model includes three basic stages: (i) the formulation of the technology assessments data set; (ii) the removal of the redundancy of the assessment criteria on the basis of the indiscernibility relations and the concept of reducts defined in the Rough Set Theory; (iii) the development of the rough variables that capture the vagueness or uncertainty and the prioritization using the SE-DEA model. The use of the hybrid model implies the consideration of two variants of the assessment: optimistic and pessimistic, formulated based on a lower and upper approximation of the rough variables. The model’s algorithms have been presented using the universal language of mathematical equations and UML schemes. An empirical application of the hybrid model to the real data of the technological foresight project «NT FOR Podlaskie 2020» Regional Strategy of Nanotechnology Development positively verified the assumed effects of its use. The advantage of the hybrid prioritisation model is that it justifies the choice of technology more rationally and objectively, simplifies interpretation and better validates the results from the perspective of decision makers. The hybrid model can be treated as a reference one in complex problems of multi-criteria comparative assessment of a large set of technologies that can be characterised using a unified, earlier defined set of attributes that describe the technology or their interrelations as well as the interactions without limitation to the specifics of the technology. The proposed model may facilitate the prioritisation of technologies in conditions of uncertainty and increase the efficiency of using private and public funds for research and development related to the implementation of future production solutions in Poland following the Industry 4.0 concept.