The multi-skilled resource-constrained project scheduling problem
A recent extension of the classic RCPSP is the multi-skilled resource-constrained project scheduling problem (MSRCPSP) in which activities require skills to be executed and the resources , i.e. the humans, possess a set of skills. The problem has been investigated in the literature under different settings, and OR&S is working intensively on this challenging problem, with results described on this page.
A new solution procedure for solving the MSRCPSP has been proposed that incorporates the concepts of depth and breadth of skills (Snauwaert and Vanhoucke, 2021). In a second study, a classification scheme for adding resource skills to the RCPSP as well as benchmark datasets are proposed (paper under submission).
- Snauwaert, J., & Vanhoucke, M. (2021). A new algorithm for resource-constrained project scheduling with breadth and depth of skills. European Journal of Operational Research, 292(1), 43–59. Download the illustrative example of this paper.
- Snauwaert, J., & Vanhoucke, M. (2022). Multi-skilled resource-constrained project scheduling problem: Overview and data generation. European Journal of Operational Research, Under Submission (2nd round).
- Snauwaert, J., & Vanhoucke, M. (2022). Mathematical formulations for project scheduling problems with categorical and hierarchical skills. Computers & Industrial Engineering, Under submission (2nd round).
We have generated four new datasets MSLIB1 to MSLIB4 to test existing and new procedures, as well as a classification scheme. A fifth set (MSLIB5) is added containing some empirical test instances
MSLIB (MSLIB.zip): Download our four libraries with one click!
- MSLIB1: A basic set with a moderately wide range of workforces and projects. The focus lies on workforce size and the number of mastered skills in the workforce.
- MSLIB2: A set with larger project instances up to 90 activities and resource skills.
- MSLIB3: A set with a large number of instances, generated with detailed settings for the resource skills. This set can be used for a detailed analysis of the impact of the workforce on the construction of a project schedule.
- MSLIB4: A set composed of the hardest instances with projects of 30 activities. The purpose of this set is to test quality of solution approaches from the literature.
- MSLIB5: This set consists of 19 empirical project instances. Their network, resource and skill structure is not well structured (empirical) but the instances can be used to validate results from the four artificial MSLIB sets into practice.
SSLIB (SSLIB.zip): Download our four libraries with one click!
A well-known extension of the MSRCPSP is the so-called software project scheduling problem (SPSP). We have not done research ourselves on this problem, but we have - based on comments by referees - extended the 4 MSLIB sets to 4 new SSLIB sets, by making the following changes:
- The project network structure for the SSLIB1 to SSLIB4 sets is identical to the MSLIB1 to MSLIB4 sets.
- The multi-skilled workforce parameters for the SSLIB1 to SSLIB4 sets is identical to the MSLIB1 to MSLIB4 sets.
- The skills requirements are binary requirements (which is different from the MSLIB sets).
- The resource assignments is based on the degree of activity dedication (which is different from the MSLIB sets).
- The activities have no fixed duration, but have now a known effort (which is different from the MSLIB sets).
- Monthly salary cost data is added to the SSLIB sets.
More information about the SSLIB set can be found in Snauwaert & Vanhoucke (2022) (under revision)
Four existing datasets from the following papers were converted to the .msrcp file format used for our MSLIB dataset:
- Correia(2012).zip: Correia, I., Lampreia-Lourenço, L., and Saldanha-da Gama, F. (2012). Project scheduling with flexible resources: formulation and inequalities. OR spectrum, 34(3):635–663.
- Montoya(2014).zip: Montoya, C., Bellenguez-Morineau, O., Pinson, E., and Rivreau, D. (2014). Branch-and-price approach for the multi-skill project scheduling problem. Optimization Letters, 8(5):1721–1734.
- iMOPSE(2015).zip: Myszkowski, P. B., Skowroński, M. E., Olech, L. P., and Oslizlo, K.(2015). Hybrid ant colony optimization in solving multi-skill resource-constrained project scheduling problem. Soft Computing, 19(12):3599–3619.
- Almeida(2016).zip: Almeida, B. F., Correia, I., and Saldanha-da Gama, F. (2016). Priority-based heuristics for the multi-skill resource constrained project scheduling problem. Expert Systems with Applications, 57:91–103.