The tool indexing problem is one of allocating tools to slots in a tool magazine so as to
minimize the tool change time in automated machining. Genetic algorithms have been suggested
in the literature to solve this problem, but the reasons behind the choice of operators for those
algorithms are unclear. In this paper we compare the performances of four common crossover
operators and four common mutation operators to find the one most suited for the problem.
Our experiments show that the choice of operators for the genetic algorithms presented in the
literature is suboptimal.