Title: Learning Local Models of Six-Axis Industrial Robot for Improved Feed-Forward Control This thesis deals with the application of data-driven identification and learning approaches when applying feed-forward velocity control to a six degree of freedom ABB industrial robot system. The work is aimed at compensation for undesired effects within the system when performing high precision tasks, such as water jet cutting. The work conducted within this thesis is mainly focused on friction compensation, since it is considered to be the largest source of error in the application.

Continued responsibility for specific weekly study sessions in the course Control Science for Computer Engineers following the academic support model "Supplemental Instruction".

Prepared a software product for open source licensing. The product is a platform for running execution graphs on the many-core chip Epiphany from Adapteva. Development included documentation and implementation of example projects for the platform utilising the many-core architecture.

Responsible for specific weekly study sessions in the course Control Science for Computer Engineers, following the academic support model Supplemental Instruction.

Integration of a Matlab-tool in Wonderware Intouch HMI including enhancements of the user experience. Work consisted of pre-study, implementation and testing. Development environment and tools included Visual Studio, Matlab, C# and Direct X. Also worked with life-time enhancement on tools for mechanical calculations made in Fortran and MS Access.

Work partly consisted of further development of a tool for sorting and analyzing events generated in Power Stations. Also partook in the development of an auto-test tool for another software tool. Programming was done in C#.

Assignment at ABB Substations in Västerås during the entire duration of the employment. Worked as a project assistant/commission engineer, performing test and modification of electrical cabinets for power stations.