NOVÁK, J. Návrh nelineárního optimálního systému řízení pro dynamicky svázané autonomní letecké prostředky [online]. Brno: Vysoké učení technické v Brně. Fakulta informačních technologií. 2024.
Ing. Jiří Novák commenced his doctoral studies under my supervision in August 2020. He excelled in all prescribed examinations in 2021 and 2022. Throughout his doctoral studies, he was actively involved in teaching the undergraduate course Signals and Systems (2020, 2021, 2022) and the graduate course Design of Cyber-Physical Systems (in English) (2023, 2024). In addition, Mr. Novák supervised three undergraduate theses and assessed three more. Overall, Jiří Novák has completed 631 credit hours of teaching, fulfilling the teaching requirements for the Ph.D. program. Since enrolling in 2020, Ing. Jiří Novák has been a valuable member of the Aeroworks research team at the Faculty of Information Technology, Brno University of Technology. As a doctoral candidate, Jiří Novák has made substantial contributions to several research projects listed below, supporting their successful completion: Framework of key enabling technologies for safe and autonomous drones' applications (ECSEL Joint Undertaking 8A19003, 826610, 2019-2023), Artificial Intelligence Driven Autonomy (TAČR TN01000029/02, 2019-2020), Laser sensor for autonomous driving of trucks (TAČR CK02000106, 2021-2023), Nanoradar for autonomous driving of trucks and its industrialization 4.0 (TAČR FW03010614, 2021-2023), Advanced compression and noise reduction for hyperspectral images data (ESA 3-17659/22/NL/GLC/mkn, 2022-2024). Jiří Novák's work on the guidance, navigation, and control components of autonomous vehicles gained global attention when it was presented nationally and internationally in connection with two new sensor systems developed within research grant projects with Valeo. Jiří Novák completed his dissertation in English, focusing on the synthesis of nonlinear optimal control systems for dynamically coupled autonomous aerial vehicles. He conducted independent research, continuously presenting credible results at major international conferences, including the International Council of the Aeronautical Sciences in 2022 and 2024; International Conference on Machine Learning, Optimization, and Data Science in 2023 and 2024; Digital Avionics System Conference in 2023 and 2024; as well as the AIAA AVIATION Forum in 2024. Additionally, Jiří Novák co-authored two software applications relevant to his dissertation: the "Mission Design and Optimization Tool," using numerical techniques to solve optimal control problems related to aerial vehicles, and a "Sensor Fusion Toolbox" focused on optimal state estimation techniques. Jiří Novák also completed a prestigious foreign internship at Technion - Israel Institute of Technology from May 2023 to July 2023, during which he co-authored a journal paper for the AIAA Journal of Aerospace Information Systems. The submitted dissertation is of exceptional quality and meets the requirements for a dissertation.
In summary, the thesis is very well written, satisfies high scientific standards, and addresses a relevant and challenging problem class. In his thesis Jiri Novak applies modern model-predictive control and estimation techniques for grasping tasks with a UAV and advances the state-of-the art with very good results. With his thesis he shows the ability to work on a high scientific level. Without any reservations I recommend to accept the thesis and to proceed with its defense. It meets the generally accepted requirements for the award of an academic degree in accordance with Section 47 of Act No. 111/1998 Coll., on higher education institution.
Mr. Novák’s contribution and work align perfectly with the current challenges in Unmanned Aerial Vehicle applications and adoption of Machine Learning algorithms in Guidance, Navigation and Control systems. His thesis demonstrates deep theoretical understanding, complemented by experience gained from working on industrial research projects. This has led to exceptional results that surpass the current state-of-the-art in Nonlinear Model Predictive Control techniques and their implementation for complex multi-degree-of-freedom aerial manipulation. Furthermore, Mr. Novák's work takes into account operational constraints, onboard optimization computational limits, and hyperparameter selection, among other factors. I grade the thesis as “Excellent” (summa cum laude) I do recommend continuation of the Ph.D. degree examination process.
eVSKP id 161902