This thesis focuses on the development of a robotic device to be used in parallel with observational learning techniques for facilitating the recovery of the upper limb in post-stroke patients. It has been shown in the existing observational learning literature that observational practice for the execution of goal-directed single arm movements can engage the mirror neuron system and motor areas involved in learning motor actions. On the other hand, robotic-based therapy protocols have proven successful in which participants are able to learn the required perception-action skill. However, robotics have not been overly successful in the generalization of learning to other tasks and this is an essential aspect on improving performance on Activities of Daily Life (ADL). Observational learning of motor skills has been shown to produce transfer across limbs and generalization across muscle groups in the same limb, as well as transfer to perceptual tasks. Therefore, our long-term hypothesis is that a combination of interactive robotics and action observation techniques might offer a greater benefit regarding transfer to ADLs in comparison to pure robotic training. The results from this research broaden the theoretical understanding of observational learning and drive the future development of rehabilitation protocols using the combination of robotic and observational learning techniques. We hypothesize that if the application of these techniques, for non-stroke individuals, yield benefits for the learning of motor/skill actions, then such paradigm will serve as a foundation in the future development of methods for facilitating the recovery of upper limb function after stroke.

The diploma thesis deals with creation of software support for the robotic device used for absolute GNSS antenna calibration, which is developed by Institute of Geodesy at Faculty of Civil Engineering, Brno University of Technology. In the beginning of this thesis, there is in general discussed issue of absolute GNSS antenna calibration. Then the text continues by description of calibration device itself together with summary of testing performed so far. Next sections are focused on its main content. There is mentioned testing calibration process, followed by all of every single steps, which leads the procedure towards gaining calibration parameters for removing errors arising due to instability of antenna´s phase center position.

The thesis is aimed at movement analysis and planning of a robotic device for absolute GNSS antenna calibration, which is developed at Brno University of Technology, Faculty of Civil Engineering, Institute of Geodesy. The first section is devoted to independent verification of the temporal synchronization of the robotic device’s movement, which is a prerequisite for the successful antenna calibration. The following section is focused on analysis of moveable characteristics since their knowledge is crucial for the following movement planning. Method of the device’s movement shooting by camera and evaluation its position based on gained photographs is used for both testing measurements. The final part is dedicated to movement planning, especially to calibrated antenna weighting influence on device’s movement characteristics, test results implementation into the observation program and arm movement planning considering the homogeneous calibration measurement layout.