This webpage contains the MP1 instances corresponding with the following paper:

Coolen, K., Wei, W., Talla Nobibon, F. and Leus, R. (2014).  Scheduling modular projects on a bottleneck resource.  Journal of Scheduling, 17(1), 67-85.  (pdf) (DOI)

The random instances from the first data set (containing only n:n-instances) can be found here.

The random instances from the second data set (containing general MP1 instances) are here.

The instances are saved as .mpo files.

Both data sets contain 10 instances for each value of the number of non-dummy jobs n = 10, 20, ..., 120 and the order strength of the induced network OS = 0.4, 0.6, 0.8.  The filename of an instance takes the form "A_nB_osC_D", where A describes the data set (A=s means first data set, A=g means second data set), B the number of non-dummy jobs, C the order strength of the induced network, and D the index of the instance in the instance group (from 1 to 10).

The first line of each instance contains two numbers: the number of jobs n+2 (including the two dummies), and  the discount rate (which is always zero for problem MP1).  Next, we have a line for each job: first we have the cost (negative for non-dummy jobs, positive for dummy-end job and equal to the payoff), followed by the probability of success and the job duration (redundant for MP1).  The other numbers are the number of immediate successors of the job, followed by an enumeration of the immediate successors.

The instances from the second data set have a slightly different format: the first line contains respectively the number of non-dummy jobs, the number of non-dummy modules, and the discount rate (zero).  Next, a line is created for each module: the line starts with the number of jobs in the module, followed by an enumeration of the jobs in that module; then the number of immediate successor modules is given, followed by the successor modules. Finally, there is a line for each job: it contains the cost of the job, the probability of success, the duration, the number of immediate successor jobs in the corresponding module and an enumeration of the successor jobs.