Publications

• Sorted by Date • Classified by Publication Type • Classified by Research Category •

MORL4Water: A Modular Multi-Objective Reinforcement Learning Toolkit for Water Resource Management

Zuzanna Osika, Roxana Rădulescu, Jazmin Zatarain-Salazar, Frans A Oliehoek, and Pradeep K Murukannaiah. MORL4Water: A Modular Multi-Objective Reinforcement Learning Toolkit for Water Resource Management. In Proceedings of the 25th International Conference on Autonomous Agents and Multiagent Systems (AAMAS), pp. , 2026.

Download

pdf [2.7MB]  

Abstract

Many real-world decision problems involve conflicting objectives.Multi-objective reinforcement learning (MORL) extends standardRL to optimize multiple objectives simultaneously, producing policysets that capture different trade-offs. However, MORL research oftenrelies on simplified benchmarks with limited real-world relevance.We present MORL4Water, a modular toolkit for creating realisticMORL environments in water resource management. Built on MO-Gymnasium, MORL4Water enables scenario construction from realdata and systematic evaluation of MORL methods. We illustrateits use on the Nile and Susquehanna rivers, benchmarking severalMORL algorithms against EMODPS, a domain-specific baseline.Beyond standard performance metrics, we analyze solution sets toreveal differences in exploration, scalability, and trade-off diversity.Our results show that most state-of-the-art MORL algorithms un-derperform relative to EMODPS, especially in higher-dimensionalsettings, and highlight the value of solution-set analysis for robust,real-world applications.

BibTeX Entry

@inproceedings{Osika26AAMAS,
    title=  {{MORL4Water}: A Modular Multi-Objective Reinforcement Learning Toolkit 
            for Water Resource Management}
    author= {Osika, Zuzanna and R{\u a}dulescu, Roxana and Zatarain-Salazar, Jazmin and 
             Oliehoek, Frans A and Murukannaiah, Pradeep K},
    booktitle = AAMAS26,
    pages={},
    year=   2026,
    organization={IFAAMAS},
    keywords =  {refereed},
    abstract =  {
Many real-world decision problems involve conflicting objectives.
Multi-objective reinforcement learning (MORL) extends standard
RL to optimize multiple objectives simultaneously, producing policy
sets that capture different trade-offs. However, MORL research often
relies on simplified benchmarks with limited real-world relevance.
We present MORL4Water, a modular toolkit for creating realistic
MORL environments in water resource management. Built on MO-
Gymnasium, MORL4Water enables scenario construction from real
data and systematic evaluation of MORL methods. We illustrate
its use on the Nile and Susquehanna rivers, benchmarking several
MORL algorithms against EMODPS, a domain-specific baseline.
Beyond standard performance metrics, we analyze solution sets to
reveal differences in exploration, scalability, and trade-off diversity.
Our results show that most state-of-the-art MORL algorithms un-
derperform relative to EMODPS, especially in higher-dimensional
settings, and highlight the value of solution-set analysis for robust,
real-world applications.
    }
}

Generated by bib2html.pl (written by Patrick Riley) on Wed Mar 04, 2026 22:15:52 UTC