Automation of metal sheet cutting increase the effectivity of the metal sheet manufacturing process. The goal of this project is object tracking algorithm that leads to optimalization of the metal sheet cutting process. The suggested object detection and tracking method is composed from multiple machine learning state of the art algorithms. The obtained results are close to the expert human operator guidelines.

The contribution presents one of some intended steps oriented to improvement of the asynchronous motor calculation model, which is as a part of the bigger simulation whole utilized within PhD study programs and SGS grants to study electromechanical phenomenon in railway vehicles traction drive. The improvement is related to modeling and integration of a pulse-width modulator and inverter for powering of the mentioned motor. As it is supposed, the ripple of the motor electromagnetic torque, which is a result of power supply from the inverter, could have an effect on torsion dynamics of the drive.

The article deals with the interaction between the machine tool and its foundation. The main problem here is the possible factual uncertainty of this interaction, especially if the machine is placed a different way than in a statically certain way. Uncertainty is meant here, in particular, the ignorance of the actually transmitted forces between the machine and its foundation, and their change in time either by the machine (effects of its own operation) or foundation (transmission of vibrations from the surrounding, or consolidation of the subsoil). The present solution of this state is to directly measure these transmitted forces between the machine and its foundation, which can be performed during machine installation and the alignment of its internal geometry or during its real operation. The author assumes that this solution, together with other necessary prerequisites, will enable to create a generalizable procedure for unambiguous and repeatable alignment of the machine geometry during its installation, to record undesirable vibrations transmitted from the surrounding to the machine, or to capture the excessive vibration of the machine. This solution will allow design more efficiently the fundament or subsoil, and in the case of the development of machine prototypes, the knowledge gained will, for example, make it possible to optimize the number of machine leveling elements and their positioning.

This paper presents an overview for many ways of controlling high-speed electrical machines, Many topics are presented, new ways and many uses for the high-speed electrical machines and theoretical study of their limits are also presented. The main goal from this paper was to determine what are the best components to build a frequency inverter that can be used to control the High-Speed Electrical machines(mainly from 0 up to 100v) in the best way with the lowest losses, The results of some experiments for testing the best components to build the frequency inverter are reported, the losses for different kinds of components that can be used to design this inverter are also reported, and finally a new kind of these components is proposed to test and use for this inverter in the future work.

This paper shows the use of the shifting method for relay feedback identification. An identification program using this method was created for the programmable logic controller Tecomat Foxtrot. The program finds two points of the Nyquist frequency characteristics from a single relay feedback test and calculates model parameters of the identified system. For this purpose it is assumed that the system is describable by the second order time delayed model. The code was written in the structured text language according to the standard IEC 61131-3. The program was tested on two simulated systems and on a laboratory controlled plant called "Air Aggregate". The obtained results show a good applicability of the proposed identification method for identification of time invariant aperiodic systems.

Understanding the biomechanics of the human body is important for applications in healthcare, sport, and also in daily life. To optimize a rehabilitation or medical surgeries, improve sport results, or improve daily life of regular people, we need to effectively collect data of their movements during their activities and daily living. Effectively means sufficient amount and suitable types of data. For that purpose, we developed various types of wearable sensors. Started with bulky system placed in a case on a belt of a patient. The system used accelerometer attached to a limb and it was connected by wires to the system. Ended with small, lightweight, and affordable system attached directly to the body, what eliminates any distractions. The system consist of 9DOF sensor and can be extended by surface EMG. The battery has enough capacity for long-term monitoring.

Since a significant and possibly major part of machine tool error is caused by thermal deformation, its compensation has high priority as an accuracy increasing tool. Various methods of compensation have been developed including a dynamic method using transfer functions. Such a method, with relatively minor calibration requirements, has proved to be reliable in the prediction of thermally induced error at one point of a machine tool workspace. This paper focuses on extending the method usability across the workspace. A possibility of extrapolating the predicted deviation of a tool in the direction of Z axis in the calibrated location to the whole workspace is examined. The concept verification on data provided by recently developed volumetric measurement method is presented. Furthermore, data obtained by volumetric measurement of kinematically different machine tools are introduced.

The automotive industry as a whole is striving to breakthrough new grounds in producing environment-friendly vehicles without compromising on performance. The stringent constraints in relation to CO2 emissions from different governments all around the world have encouraged OEM’s to adopt different solutions. Due to which in recent years a trend of cutting out weights of different sub-systems in the vehicles has been observed to be one among them. In this paper, the possibility of reducing the weight of the seating systems through a modular actuator system is discussed. On observing the seating systems as a whole it could be noted that the seating system as a whole has not changed much. On comparing the weights of an entry level manually actuated seating system with a high-end luxury motorized seating system, the difference in weight could be almost 100% in certain cases. The demand for luxury in seating systems gave way for motor powered seating systems. The OEM’s kept on adding more and more actuators, wiring harnesses and control systems for every manually operated actuation rather than redesigning the seating structure owing to time and cost effects. This trend resulted in the seating systems to be one of the most complicated systems in a vehicle. Thus this paper will discuss a possible solution towards cutting down the weight of the seating system as a whole by introducing a modular actuator design.

In this study, I deal with a description and evaluation of remodeling changes after implantation of the short stems. First postoperative image was used to create 2D model of proximal femur with the implant as a geometry for a boundary value problem describing mechanical interaction between bone tissue and hip stem. The problem was formulated by means of finite element method. To evaluate remodeling changes, FEM results were compared to two- years postoperative image. The comparison was based on changes in stress field and changes in principal stress directions. The results suggest, that short stem Proxima helps to restore the orientation of the main groups of trabeculae in the proximal femoral area and reduces the stress shielding effect after THA.