The ever-increasing demand for computing has led to the need for specialized heterogeneous hardware, and the frameworks required to utilize them. Besides the traditional central processing units, more and more programs will make use of specialized hardware to accelerate computations. However, the increase in computing also leads to shorter mean time between failures. In this thesis, we apply fault tolerance to Portable Computing Language (PoCL), an open-source implementation of the OpenCL standard. We show that our solution is easy to apply to existing programs making use of PoCL/OpenCL and is able to greatly reduce the total number of errors visible to the end user. Our solution can be used on any device supported by PoCL and provides a low overhead, given that the hardware requirements are met.

Grading exams is a time-consuming activity for teachers. Zesje is an open-source tool created to aid teach-ers in exam grading and streamline the grading process. Zesje currently uses computer vision techniques torealign images, and automatically find student numbers. However, teachers can currently only use Zesje tograde questions manually. Moreover the computer vision capabilities of Zesje can be improved. To make iteasier to grade exams, it should be possible for teachers to have multiple choice questions graded automati-cally. This project describes various improvements for Zesje, most notably using computer vision for the auto-matic grading of multiple choice questions, improving the accuracy of aligning scanned submissions, andautomatically detecting blank solutions. The team had to make several choices regarding implementations and choice of technology. Design goalswere also created to serve as a guideline for the project. At the end of the project, with the features imple-mented by the team, Zesje can automatically grade multiple choice questions, identify blank solutions andhas the corresponding front-end changes that allow the user to create multiple choice checkboxes on theexam PDF. These features have been tested extensively. The use of Zesje also poses some ethical challenges. Using automated grading may result in the event thatsome submissions may never be seen by a grader. By using benchmarks to compare the performance of processing scans in Zesje, the team found out thatthe grading time has greatly been reduced.