Job assignment is a well-known problem in manufacturing lines. Line management tries to maximize line productivity by assigning a given crew a set of predefined line jobs. However, problems may arise because of abnormal conditions such as absenteeism, workstation failure, parts shortage, or unevenly distributed work-in-process. This report considers a dynamic job-assignment policy making use of online information for assignment. This policy attempts to achieve a good skill match between jobs and operators, and to eliminate unnecessary assignments. To fulfill this policy, a two-part assignment program is proposed: an algorithm for job matching and a set ofad hocrules. The matching algorithm is well established in literature, but thead hocrules are application-dependent. Discussion here concentrates on thead hocrules and online data collection.