HeuristicLab 3.0

HeuristicLab 3.0 can be said to combine the intuitive GUI of HL1.1 with the modularity of HL2.0 to a new Otimization Environment that is more powerful than ever and one that we want to push to a public release.

During Summer 2007 and after the experience we made with HL1.1 and HL2.0, Stefan kicked off the idea of a new major version that should be built from the ground up with an improved user interface and the strengths and weakness of our previous architectures in mind. The main requirements of this new optimization environment are threefold: To be as general as possible and yet as best fitting for heuristic optimization tasks, to expand the modularity that we introduced with HL2.0 even further and to be scalable to parallel computer systems and complex Algorithms. Quite impressingly half a year later this software architecture has formed shape, a new face and has established as our new base for use in teaching and industry projects.

The genericity of HL3.0 allows to build almost any kind of algorithm through a click-and-drag user interface. The operators that form the algorithms are specifically targeted for the heuristic optimization community. The algorithms that are created through the combination of these operators in the UI live in the context of an engine which decides upon the execution.
The modularity of HL3.0 allows to combine any number of operators which can be saved and reused. The algorithms that are created in HL3.0 are saved in files called workbenches which can be distributed and modified. Thus algorithm development does not happen in a programming language anymore and parts of the algorithms can be shared among workbenches.
The scalability of HL3.0 allows to fully use a Dual-Core machine or even a multicomputer parallel system such as a cluster or GRID system making use of today's increased parallel processing power.

Screenshots

Below are several screenshots of HeuristicLab 3.0. For a larger version please click on the images.