Browsing by Author "Nieuwenhuizen, Florian"

Now showing 1 - 2 of 2

Open Access
Evaluation of typing performance in virtual reality on a physical keyboard
(2017) Nieuwenhuizen, Florian
Nowadays, when people talk about virtual reality (VR), most think about playing computer games. VR has many fields of use, like for training and at work, that most people haven’t thought about. Input, especially text input has always been difficult for VR. Many different attempts at allowing text input in VR have been made, but none have ever combined the most used input method for the computer, the keyboard, with sub-millimeter accurate, visible hands. In the scope of this thesis, a prototype combining a motion capturing system, OptiTrack, with VR technology and the Unity game engine, was created to allow viewing hands in realtime, while inside of a virtual environment. Using this prototype, we wanted to find out, which effect hands have on typing inside of a virtual environment. Furthermore, we were interested, if there are differences in typing performance while typing inside of VR in comparison to typing outside of VR. Users will be typing while viewing different forms of hands. Is typing performance negatively impacted if no hands are visible? To answer this, different hand models were created for performing a typing test while immersed inside of the virtual environment. A study 32 participants was conducted, where each participant completed the typing test seven times while in VR and once outside of VR. Each VR condition contained a different hand model. These participants were divided into two equal typing groups, a slow and a fast one. On average, when analyzing the data of all participants at once, we found no statistically significant difference in words per minute (WPM) or accuracy when comparing typing inside of VR to typing outside of VR. The slow typers, did show a statistically significant difference in WPM and accuracy in this comparison. Furthermore, it was discovered, that seeing any sort of hand model while typing in VR is important. Not being able to see hands had a negative effect on typing performance, increased task load, and also decreased the perceived presence in VR. We found that using different hand models inside of VR does not statistically significantly affect typing performance, as long as some variation of a hand is visible.
Open Access
Time-sensitive Deployment and Management for Cyber-Physical Systems using TOSCA
(2020) Nieuwenhuizen, Florian
When managing systems, it is often difficult to find a suitable point in time to update without disrupting the availability. As Cyber-Physical Systems influence the physical environment, the importance of safely managing such systems, in order to avoid dangers, increases. Context-aware systems have previously been configured to only execute management processes when a set of situations is valid. Here, a situation can be understood as any event capable of changing, often connected with some form of sensor. This enables improved timing, but could still lead to an undesirable situation change during an update, resulting in a decrease of availability or forcing a rollback. To avoid situation changes, this thesis proposes executing processes in a time-frame during which all situations stay in their correct state. Scheduling the execution of a management operation during a time-frame is only possible when the duration of the process is known. Using TOSCA and the OpenTOSCA Runtime a prototypical implementation allowing the measurement of management processes in order to generate a worst-case execution time is introduced. Generating timing information enables context-aware systems to estimate if an operation can be performed during a time-frame, therefore decreasing the likelihood of executing management processes at undesired times. The timing information from plans can then be used to create profiles for often occurring actions, enabling the possibility of providing users with timing information while modeling the topology of a system.