After two years of experimenting with various methods and algorithms, from brute force methodologies, through models of evolving life algorithms, to genetic algorithms with dynamic DNA structures, I could not get a method that worked well and consistently. Most of these methods worked well for simpler two dimensional type schedules, e.g. night shift, sport roster or bus timetables, but seemed to fall over when working with the multi dimensional concepts of the school timetables.
I eventually chucked all of these ideas out, sat down and devised a straightforward logical method, also based on nature, what I like to call Hierarchical Contention, or in simpler terms, the pecking order, be it a pride of lions, a troop of monkeys, or even the set of groups used in timetables. The engine emulates Mother Nature so closely that if you could see the battles that take place within the program, I would have to rate TimeDesign PG13 for violence.
Using this method, solving times have reduced dramatically, where the average South African high school timetables can usually be solved within seconds on today's entry level computers. Having played around with similar programs, I can quite confidently say that the solving engine is one of the best and fastest available.