prof. Ing. Martin Hartl, Ph.D.

1. Adaptive control design be means of the ROS framework

   A significant development in the area of mobile robotics was noticed in the last time in connection with the arrival of modern, powerful and mainly miniature computers and controllers. The development of mobile robotic devices is also correlative of to the development of sensory systems that are increasingly more accurate, faster and smaller. The role of mobile robots can be different. In some stores, Amazon, for example, a system of autonomous mobile robots is used. These mobile robots are able to bring and carry off cart with the desired goods which is needed. Furthermore, mobile robots are used for example for mapping indoor or outdoor space, such as service robots, or as a part of other mechanism, such as parking assistant in automobiles. The proposed project is focuses on mapping the interior and aims to design and implement algorithms of adaptive control of mobile robots. The project will design a system for obtaining data to create a 3D model of the mapped area. For designing the algorithm and for robot control the ROS framework will be used.

   Tutor: Šťastný Jiří, prof. RNDr. Ing., CSc.

2. Analysis and identification of local DNA structures by means of soft computing methods.

   The project contains the develop of a specialized platform for DNA sequence analysis with focus on large volumes of data that will include required algorithms for identification of structures like triplexes, quadruplexes and for protein motifs analyses with reporting and visualization tools and their storage in suitable database system. Proposed software will be implemented as a web service and will be available online for public usage. The software will be used for characterization and evaluation of the local DNA structures in DNA sequences with the focus to possibility to analyse whole genomes and different local DNA structures including triplexes and quadruplexes.

   Tutor: Šťastný Jiří, prof. RNDr. Ing., CSc.

3. Analysis and optimization of business data with the use of artificial intelligence

   The aplication of artificial intelligence methods for data optimization is especially suitable for problems which are very difficult to solve using classical mathematical methods or where using deterministic computation methods would require unacceptable simplification of the problem. An acquisition of the optimization data by means of modern methods of artificial intelligence is a significant task in object classification and prediction. Approaches used by a graduant solving the aforementioned task are as follows: Analysis artificial intelligence methods, analysis algorithms for data classification and prediction.

   Tutor: Šťastný Jiří, prof. RNDr. Ing., CSc.

4. Distributed optimazation systems

   The distributed approach is the current trend in many areas of computer applications, in communications, databases, calculations, control applications, and many others. For all these areas, optimization is an important part of their implementation. Distributed Optimization Approach allows to customize the structure of solved optimization tasks to real-world conditions where sub-optimization can be performed at lower levels to interact with other sub-optimizations that will contribute to the overall result.

   Tutor: Roupec Jan, doc. Ing., Ph.D.

5. Holographic visualization methods for industrial applications

   Industrial visualizations are almost llimited to 2D imaging technologies and visualization techniques. Industrial IoT systems generate huge data sets, which must be evaluated using AI-ML and intuitive visualizations. Holographic systems provide far more possibilities for intuitive visualization for big data problematic. Holografic methods and systems differ in parameters and suitability for industrial applications. There is a need to select and propose methods of holographic visualization suitable for different kinds of industrial appplications.

   Tutor: Roupec Jan, doc. Ing., Ph.D.

6. Intelligent Model Predictive Controllers

   Model predictive control (MPC) is an efficient method for the controller design of a large number of processes. However, linear MPC is often inappropriate for controlling nonlinear large-scale systems, while non-linear MPC is, in general, computationally costly. The work will try to find intelligent and optimal design of MPC for controlling of selected nonlinear and complex systems. Research and applications in the areas of advanced control theory, computational intelligence, simulation and practical real-time implementation will be challenges of this work.

   Tutor: Matoušek Radomil, prof. Ing., Ph.D.

7. Nontraditional methods for data classification and prediction in process control

   Artificial intelligence methods belong among modern methods for optimization of process control. Using theese methods is especially suitable for problems which are very difficult to solve using classical mathematical methods or where using deterministic computation methods would require unacceptable simplification of the problem. An acquisition of the optimization data by means of modern methods and algorithms of artificial intelligence is a significant task in object classification. Approaches used by a graduant solving the aforementioned task are as follows: Analysis artificial intelligence methods, analysis algorithms for data classification and prediction.

   Tutor: Šťastný Jiří, prof. RNDr. Ing., CSc.

8. Optimisation of Serviceability in Network Applications

   In applications that serve locations deployed in a large area for certain customer service, it is a typical task to minimise these locations so that each customer has at least one of the centers at the available distance. The problem of coverage for this task has O (2 ^ n) complexity, where n is the number of given places and it is necessary to solve it by heuristic methods for the "large" instances of the problem. However, the task has even more complex formulations considering service capacities and customer requirements. In the dissertation the aim is to apply a general problem solving in the problems of communication of 5G mobile networks and data storage in NoSQL databases.

   Tutor: Šeda Miloš, prof. RNDr. Ing., Ph.D.

9. Robot Motion Planning in a Scene with Obstacles

   Methods for robot motion planning have a number of applications, they can be used in production halls where the robotic device moves on a defined route, but also in situations where it is necessary to construct the track continuously, e.g., when searching for explosive areas, searching for persons in emergency situations (e.g. earthquakes). Traditional approaches include scene decomposition methods, potential field method, and roadmap methods. Within a group, a number of specific methods can be defined, which may vary, for example, according to the geometric structures used (visibility graphs, rapidly-exploring random trees, Voronoi diagrams). The task is to classify and compare main approaches and implement an algorithm to ensure that the robot path is smooth and safe from the point of view of the threat of collision with obstacles.

   Tutor: Šeda Miloš, prof. RNDr. Ing., Ph.D.

10. Robotics and Bin Picking - The Search for the Holy Grail

    There are three main types of bin picking: structured, semi-structured, and random bin picking. Each presents an increasing level of application complexity, cost and cycle time. We can say that random bin picking is already approaching mainstream of robotics. Random bin picking requires a convergence of technologies, particularly three main components that raise the robot’s intelligence: sensors, software, and end-of-arm tooling. Development in all three areas is moving us ever closer to that elusive prize.

    Tutor: Matoušek Radomil, prof. Ing., Ph.D.

11. Stabilisation and Control of dynamical systems exhibiting a chaotic behavior

    Some dynamical systems exhibit a complex behavior known as deterministic chaos. The topic is focused on analysis and control of suitable chaotic models (with respect to a widest set of system's parameters) using computational intelligence.

    Tutor: Matoušek Radomil, prof. Ing., Ph.D.