Project Data
Welcome to the OR&S project database
The Operations Research & Scheduling research group has been generating, collecting, and manipulating project data for the resource-constrained project scheduling problem for over 20 years. The story of this fascinating research journey is summarized in the book "A Quest for Projects with Scarce Resources". All project data are available for download on this website.
On this page, you will find different databases for project portfolio scheduling, risk analysis and project control.
- Summary of the datasets: A short summary of this webpage.
- Artificial project data: Several datasets containing +30,000 instances (and the option to create up to almost 4 million project instances)
- Empirical project data: 1 dataset with real projects with +180 project instances
- Project portfolio data: 2 datasets with almost 40,000 project portfolio instances
- Solutions Update: A new website where you can download and upload best known solutions
- Bookstore: Get to know how to use data for managing projects
- Quick links: If you don't want to read everything on this website, go to the bottom of this page and click on the quick links.
If you would like to know more about data and anything else related to our project management research, visit www.or-as.be.
1. Summary of the datasets
Quick Overview by Problem Type
| Problem Type | Main Dataset(s) |
|---|---|
| RCPSP (single-project scheduling) | RG30, RG300, DC1, DC2, MT, PSPLIB, CV, sD, MH, NetRes, 1kNetRes, VNR, LPP, LPSP |
| MMRCPSP (multi-mode) | MMLIB50, MMLIB100, MMLIB+ |
| RCPSP-AS (alternative subgraphs) | ASLIB |
| MSRCPSP (multi-skilled) | MSLIB, SSLIB |
| Dynamic Scheduling (real projects) | DSLIB |
| RCMPSP (Multi-project scheduling) | MPLIB1, MPLIB2 |
Project Data Libraries
| Category | Dataset(s) | Description | #Instances | Reference |
|---|---|---|---|---|
| Artificial Resource-Constrained Data (RCPSP) |
RG30 | Benchmark RCPSP datasets | 1,800 | Vanhoucke et al. (2008) |
| RG300 | Large-sized RCPSP instances | 480 | Debels & Vanhoucke (2007) | |
| DC1 and DC2 | Benchmark RCPSP datasets (used with cash flows) | 1,800 + 720 | Vanhoucke et al. (2001) (DC1), Vanhoucke (2010) (DC2) | |
| MT | Diverse datasets (no resources) | 4,100 | Vanhoucke (2010) | |
| PSPLIB | Classic RCPSP benchmark library | 2,040 | Kolisch & Sprecher (1997) | |
| CV | Hard RCPSP instances for exact solution methods | 623 | Coelho & Vanhoucke (2020) | |
| sD | Hard RCPSP instances for heuristic methods | 390 | Coelho & Vanhoucke (2023) | |
| MH | RCPSP instances for matheuristic approaches | 461 | Vanhoucke & Coelho (2024) | |
| NetRes | Large-scale RCPSP dataset combining network and resource data | ~4 million | Vanhoucke & Coelho (2018) | |
| 1kNetRes | Subset of NetRes | 3,810 | Derived from NetRes | |
| VNR | Subset of NetRes | 1,750 | Van Eynde et al. (2024) | |
| LPP, LPSP | Large-sized RCPSP instances | 161 | Luo et al. (2022) | |
| Instance Variants | Original, lowRU, highRD, and optimal-equivalent versions | 4 versions per instance | Vanhoucke & Coelho (2021, 2024) | |
| Artificial Multi-Mode Data (MMRCPSP) | MMLIB50, MMLIB100, MMLIB+ | Multi-mode RCPSP datasets | 4,320 | Van Peteghem & Vanhoucke (2014) |
| Artificial Alternative Subgraph Data (RCPSP-AS) | ASLIB | RCPSP with alternative subgraphs | 115,920 | Servranckx & Vanhoucke (2019) |
| Artificial Skilled Resource Data (MSRCPSP) | MSLIB | Multi-skilled RCPSP datasets | 32,480 | Snauwaert & Vanhoucke (2023) |
| Empirical Project Data | DSLIB | Real-life projects with scheduling, risk, and project control data (Dynamic Scheduling) | 200+ projects (originally 52) | Batselier & Vanhoucke (2015) |
| Project Portfolio Data (RCMPSP) | MPLIB1, MPLIB2 | Resource-constrained multi-project scheduling datasets (RCMPSP) | 39,635 | Van Eynde & Vanhoucke (2020, 2021) |
2. Artificial project data
OR&S has generated several datasets with several project instances for different versions of the well-known resource-constrained project scheduling problem.
The resource-constrained project scheduling problem (RCPSP)
- Download our 13 datasets (from GitHub). This folder contains an excel file with best known solutions, and the following datasets:
- A summary of five datasets (RG30, RG300, DC1, DC2 and MT) is given in Vanhoucke et al. (2016). Each file comes from a different source, i.e. Vanhoucke et al. (2008) (RG30), Debels and Vanhoucke (2007) (RG300), Vanhoucke et al. (2001) (DC1), Vanhoucke (2010) (DC2) and Vanhoucke (2010) (MT)
- The PSPLIB instances are proposed in Kolisch and Sprecher (1996).
- A dataset CV with 623 hard instances (for exact procedures) is proposed in Coelho and Vanhoucke (2020).
- A dataset sD with 390 hard instances (for heuristic procedures) is proposed by Coelho and Vanhoucke (2023).
- A new dataset MH with 461 instances (for matheuristic procedures) is proposed by Vanhoucke and Coelho (2024).
- The full NetRes dataset can be generated using the MT set (only network data) and the NetRes data files (only resource data) containing almost 4 million projects as proposed in Vanhoucke and Coelho (2018).
- The 1kNetRes is a subset of NetRes and contains 3,810 instances.
- The VNR set is another subset of NetRes with 1,750 instances proposed in Van Eynde et al. (2024).
- The LPP and LPSP datasets are large-sized instances proposed in Luo et al. (2022).
- Each instance consists in four versions: the original instance, the lowRU and highRD instances (Vanhoucke and Coelho (2021)) and the optimal equivalent instances (Vanhoucke and Coelho (2024)).
- Visit RCPSP webpage for further details.
The multi-mode resource-constrained project scheduling problem (MMRCPSP)
- Download the MMLIB50, MMLIB100 and MMLIB+ datasets (from GitHub).
- Visit our MMRCPSP webpage for details.
- Read Van Peteghem and Vanhoucke (2014).
The resource-constrained project scheduling problem with alternative subgraphs (RCPSP-AS)
- Download the ASLIB dataset (ASLIB.zip).
- Visit our RCPSP-AS webpage for details.
- Read Servranckx and Vanhoucke (2019).
The multi-skilled resource-constrained project scheduling problem (MSRCPSP)
- Download the MSLIB dataset (from GitHub) and the SSLIB dataset (SSLIB.zip). The MSLIB file contains an excel file with best known solutions.
- Visit our MSRCPSP webpage for details.
- Read Snauwaert and Vanhoucke (2023).
3. Empirical project data
OR&S has also created a project dataset with project instances from companies, initially consisting of 52 projects, but now grown to +180 real projects. This dataset contains project networks with resources for scheduling the project (RCPSP) but also risk data and project control data. The integration of project scheduling, schedule risk analysis and project control is called "Dynamic Scheduling" (DS, Vanhoucke (2013)).
- Download the DSLIB dataset (from GitHub).
- Visit our empirical project data webpage for details.
- Read Batselier and Vanhoucke (2015) (this paper presents the first 52 projects) and the "The Illusion of Control" (updated to 181 projects)
4. Project portfolio data
The resource-constrained multi-project project scheduling problem (RCMPSP)
- Download the MPLIB1 and MPLIB2 datasets (from GitHub).
- Visit our multi-project webpage for details.
- Read the papers by Van Eynde and Vanhoucke (2020) and Van Eynde and Vanhoucke (2021).
5. Solutions Update (download data and upload solutions)
We have developed a new tool "SolutionsUpdate" to download and upload solutions for the resource-constrained project scheduling problem using all datasets described above. The system is explained in Vanhoucke and Coelho (2018) and the new tool and website can be accessed on http://solutionsupdate.ugent.be.
6. Bookstore
Visit our bookstore and get to know how to manage projects with data. The bookstore contains 3 free books and 9 published books:
- Data-driven project management with Python: Optimizing schedules, simulating risk and analyzing project performance through 10 example experiments (2026, Springer)
- Data-driven project management: The science & stories behind project decisions (2026, Lannoo)
- A quest for projects with scarce resources: Seeking schedule intelligence through project data discovery (2024, Springer)
- The illusion of control: Project data, computer algorithms and human intuition for project management and control (2023, Springer)
- The data-driven project manager: A Statistical Battle Against Project Obstacles (2018, Apress)
- Integrated project management sourcebook: A technical guide (2016, Springer)
- Integrated project management and control: First comes the theory, then the practice (2014, Springer)
- Project management with dynamic scheduling (2013, Springer)
- Measuring time: Improving project performance using earned value management (2010, Springer)
7. Quick links
