A number of possible, significant extensions to this study could
be investigated in future research. It would be interesting to
consider the problem with sequence-dependent setup times (or
costs). Since the jobs are released in different points in time, in such
a situation two versions of the problem may then be investigated:
setup can be assumed to begin either when both the corresponding
job and the machine are available, or when the machine is available
even when the job is not released. Another extension would be to
consider the case in which there are precedence constraints
between the jobs belonging to the same customer, or even between
all jobs. Moreover, it would be interesting to study the problem
where unavailability constraints are imposed on the machine, i.e.,
when it is not continuously available for processing, for example,
due to preventive maintenance.