A job scheduling and rejection problem considering self-contained and cross-functional jobs

\nIn today’s large projects and complex assembly lines, a single multi-skilled worker often needs to complete jobs requiring multiple areas of expertise. Even if a worker possesses all necessary skills, their proficiency can vary. This variability makes it challenging to assess a job’s cost-performance ratio before assignment. Larger problem sizes often involve many such jobs that need scheduling or rejection. Clearly, the processing times and workers in the presented problem are more complex than traditional scheduling problems with single-valued processing times and single-functional machines. Two important observations serve as the motivation. First, traditional genetic algorithms with fixed-length chromosomes may not effectively handle the complexity of self-contained and cross-functional jobs and multi-skilled workers. Second, traditional genetic algorithms cannot guarantee a certain level of solution quality. Motivated by these observations, a novel genetic algorithm is developed. This algorithm can quickly search the solution space using an outbreeding technique. Additionally, an upper bound is provided to ensure solution quality. Experimental results demonstrate that the proposed genetic algorithm is superior to others through comprehensive comparisons.\n