The document proposes genetic algorithm approaches to solve the job shop scheduling problem (JSSP). It introduces new crossover and mutation operators designed based on the problem characteristics. The crossover combines operation orders of different machines from two parents. The mutation permutes successive operations on the same machine that are part of the critical path. Experimental results test the approach over 10 runs with a population size of 100, crossover probability of 0.7 and mutation probability of 0.1.