Keynote Speakers


Colonel Andersonn Kohl, Brazilian Army

“On the suitability of variability modeling upon the Brazilian Army Family of C2 Applications”

Abstract: this presentation will explain some C2 generic concepts, the C2 software development within the Brazilian Army aligned to the doctrine and present some points that will address the need for variability modeling in order to deliver the right products for the different army missions needs.

Bio: Col Kohl has been in the military since 1987, as a communication officer. He graduated as telecommunication engineer at Military Engineering Institute in 1996. Throughout his career he was assigned to join the SIVAM Project team, under the Brazilian Air Force management; to be the project manager of Brazilian Army command and control software development and the project manager of ground segment of the Brazilian Ministry of Defense Tactical Data Link System Project. Since 2012 he has been assigned as chief of the Command and Control Division at the Brazilian Army Systems Development Center. He has strong interest in system engineering (SE) and system of systems engineering (SOSE).

Bruce Watson

Prof. Dr. Dr. Bruce Watson, Stellenbosch University, South Africa

“Taxonomy-based Software Construction for Algorithmic Families”

Abstract: Taxonomy-based software construction (TABASCO) is a methodology for building toolkits/libraries for families of algorithms. After selecting a specific algorithmic field (for example sorting, compression, pattern matching, etc.), the methodology proceeds with constructing a taxonomy of the known algorithms for the problem. In principle, the algorithms have the same postcondition, though in practice there are some minor variations, for example co-taxonomizing single-keyword and regular expression pattern matching algorithms. The taxonomy is built using correctness-preserving refinements, starting with a generic (usually unimplementable) algorithm at the root, and arriving at all of the known algorithms, and occasionally some entirely new algorithms. The taxonomy tree highlights commonalities and variations between the algorithms. The taxonomy structure is then mapped to the software structures (class hierarchy or shared sub procedures) which emphasise code sharing. The resulting toolkit displays a more coherent structure than ad hoc toolkits, as well as better correctness, thanks to the correctness-by-construction refinement technique. This talk will include the discussion of one of the several algorithmic areas to which TABASCO has already been applied.

Bio: Professor Bruce Watson is co-founder and director of the FASTAR Research Group and the Centre for Decision Making and Knowledge Dynamics, both at Stellenbosch University. In addition, he holds a visiting professorship in Computer Science at King’s College London, and is visiting scientist at IP Blox. In 1995, Watson received his first Ph.D in computing science from Eindhoven University of Technology, after studying discrete mathematics and computer science at the University of Waterloo. He later returned to Eindhoven as chair of Software Construction. Watson’s second Ph.D, in computer science, is from the University of Pretoria in 2012. His recently co-authored book is The Correctness by Construction Approach to Programming, 2012. Parallel to his academic career, he worked as a compiler engineer at several companies (e.g. Microsoft), followed by engineering and architecture work on virtual machines and pattern matching algorithms (e.g. for Cisco). Watson’s first dissertation contained taxonomies and new algorithms for pattern matching and regular expressions. With more than a dozen researchers, his FASTAR Research Group performs fundamental research in new algorithms and implementations for high-performance pattern matching. IP Blox develops leading edge technologies for deep-packet inspection, with a focus on performance and virtualization.