Aug 2026· Systems· Vol 14, pp. 911· 0 citations· 45 references
TL;DR
A rollout-based approximate dynamic programming (ADP) algorithm that approximates it via Monte Carlo simulation embedded with a randomized base policy and restricts the action space to ten representative allocation policies to minimizes the total human resource cost.
Abstract
Knowledge-intensive projects in research and development (R&D), software, and high-technology sectors are delivered by knowledge workers who each command several skills. When many such projects compete for a shared pool of multi-skilled workers, and neither project arrival times nor durations are known in advance, deciding who works on what at each moment becomes a sequential decision problem. Therefore, we propose the multi-skilled, multi-project dynamic scheduling problem with random project arrivals and uncertain durations and formulate it as a Markov decision process (MDP) that minimizes the total human resource cost. Since the cost-to-go function is computationally intractable, we develop a rollout-based approximate dynamic programming (ADP) algorithm that approximates it via Monte Carlo simulation embedded with a randomized base policy and restricts the action space to ten representative allocation policies. On benchmarks extended from the Project Scheduling Problem Library (PSPLIB), the proposed policy lowers the average total cost by 3.1% to 16.7% relative to Q-learning on medium- and large-scale instances while completing more projects with shorter delays; on small-scale instances Q-learning attains a lower nominal cost, revealing a trade-off among cost, completion rate, and delay.
This paper aims to propose a discrete-event simulation framework for solving the Stochastic Network-based Resource-Constrained Project Scheduling Problem (SN-RCPSP), which involves uncertainty in both activity duration and occurrence, and allows for feedback loops and limited renewable resources.
A novel simulation methodology is implemented using Enterprise Dynamics software. The model integrates probabilistic branching, stochastic durations and resource constraints in an Agent-Based Modeling architecture. The approach is validated using benchmark cases, including statistical and analytical cross-validation against known methods.
The proposed model achieves comparable or superior accuracy to existing simulation and analytical methods with significantly fewer replications. It also allows for performance evaluation under different resource availability scenarios and provides managerial insights into optimal resource allocation.
This study models only one type of renewable resource and assumes fixed execution modes for activities. Learning effects and cost structures are not included, which can be explored in future research.
The simulation model offers a robust tool for practitioners in project planning, especially in uncertain and dynamic environments like pharmaceutical approvals or R&D projects. It allows managers to assess risks and resource impacts before execution.
This paper contributes a comprehensive and flexible simulation-based methodology for SN-RCPSPs, validated via empirical, theoretical and scenario-based analyses. It extends classical models by incorporating resource constraints and supports performance-driven decision-making in stochastic project environments.
A. Rahimifard, Isa Nakhai, K. Khalili-Damghani et al.· Journal of Modelling in Mana...· 0 citations
PGMPO is proposed, a novel learning framework consisting of a simple but effective multi-policy modeling approach that allows a single network to represent multiple decision-makers, and a preference-driven model optimization method that effectively guides policies to learn diverse and specialized problem-solving strategies without the need for explicit reward functions.
Inguk Choi, Woo-Jin Shin, Sang-Hyun Cho et al.· 0 citations
Project scheduling faces high uncertainty due to fluctuating productivity, weather disruptions, resource shortages, and rework. Existing models such as CPM, PERT, and deterministic optimization lack the capability to dynamically adapt to real-time disturbances. This paper proposes a novel Adaptive Stochastic Resource-Constrained Scheduling Model (ASRCSM) that integrates (i) stochastic task durations, (ii) dynamic resource allocation rules, and (iii) an adaptive re-scheduling mechanism triggered by project state deviations. In addition, this study proposes a novel Adaptive Hybrid Genetic Algorithm (AHGA) which combines Serial Schedule Generation Scheme and local search heuristics to solve the problem. The approach provides a practical, computationally efficient tool for construction planners facing uncertainty.
Nhat-To Huynh· Journal of Science & Technol...· 0 citations
Sequential decision-making in real-world applications often involves uncertainty about the environment's model. Uncertain Markov decision processes (UMDPs) represent the possible environments as a set of MDPs with shared states and actions but potentially different transition probabilities and rewards. Optimizing a single policy across all possible MDPs may sacrifice performance, while preparing an individually optimized policy for every MDP may violate operational, regulatory, or interpretability constraints on the number of policies that can be prepared and deployed. We consider settings in which model uncertainty is resolved shortly before execution, allowing the most suitable policy to be selected from a limited set prepared in advance. We introduce $k$-adaptable policy synthesis, which optimizes such a set of $k$ policies under a minimax-regret objective. We prove that the problem is NP-hard and develop KAPS, an exact nested branch-and-bound algorithm with problem-specific bounds and heuristics. KAPS jointly optimizes which MDPs share a policy and the policies themselves. Experiments across various UMDP benchmarks show that the largest reduction in regret consistently occurs when increasing from one to two policies. In the single-policy setting, KAPS is competitive with existing methods in solution quality and proves optimality substantially more often.
In engineering R&D organizations, resource contention among autonomous projects creates complex scheduling challenges in distributed multi-project environments. Traditional coordination mechanisms often prioritize global efficiency while overlooking inter-project fairness, which can lead to stakeholder resistance and execution delays. This study proposes a transparent decision support approach for the Distributed Resource-Constrained Multi-Project Scheduling Problem (DRCMPSP). We develop a two-stage scheduling mechanism that integrates Multi-Criteria Decision-Making (MCDM) into the global coordination process. First, an enhanced Chaotic Genetic Algorithm (CGA) with elitism generates local schedules. Second, global resource conflicts are resolved using MCDM methods. Inter-project fairness is implemented by limiting each project’s relative objective deterioration and by evaluating the dispersion of the resulting project-level burdens. Validation through an Unmanned Aerial Vehicle R&D case study and extensive experiments shows that the TOPSIS-based mechanism achieves a significantly lower standard deviation than the auction-based mechanism under high resource contention. The approach supports transparent and fairness-aware conflict resolution by making trade-offs among project-level outcomes explicit to managers.
Zheng Yang, Xiao-kang Wang, Jian-qiang Wang et al.· Symmetry· 0 citations
Algorithmic team formation is a well-studied problem in workforce analytics. Most prior research has focused on team-oriented outcomes, such as cost and performance, while individual-oriented outcomes have been limited to operational constraints. We formally define and solve a new variant, the growth-aware team formation problem (GATFP), which forms teams to maximize individual skill growth. Our formulation lets organizations define their own notion of growth and their own policy for allocating growth opportunities, and integrates standard team formation constraints such as covering a portfolio of projects, capping projects per individual, and meeting each project’s skill, level, and coverage requirements. We prove the problem NP-hard and develop a family of metaheuristics: randomized and greedy construction, a ruin-and-recreate large neighborhood search (RR), and a hybrid evolutionary algorithm (HE). Where the objective is linear, a mathematical program provides an exact baseline. We evaluate them under an equal runtime budget on a synthetic benchmark and a real-world-derived dataset. Both metaheuristics surpass the construction baselines and are closely matched, and which one leads depends on the growth model: RR is attractive for its solution quality and implementation simplicity, whereas HE is the stronger choice when an individual’s growth depends on their team’s composition.
Alexandros Baltas, Theodoros Lappas, Panagiotis P. Repoussis· Algorithms· 0 citations