Publications Neural Network Control
Closing the Intent-to-Reality Gap via Fulfillment Priority Logic
IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2025), 2025, Hangzhou, China
Abstract
Practitioners designing reinforcement learning policies face a fundamental challenge: translating intended behavioral objectives into representative reward functions. This challenge stems from behavioral intent requiring simultaneous achievement of multiple competing objectives, typically addressed through labor-intensive linear reward composition that yields brittle results. Consider the ubiquitous robotics scenario where performance maximization directly conflicts with energy conservation. Such competitive dynamics are resistant to simple linear reward combinations. In this paper, we present the concept of objective fulfillment upon which we build Fulfillment Priority Logic (FPL). FPL allows practitioners to define logical formulae representing their intentions and priorities within multi-objective reinforcement learning. Our novel Balanced Policy Gradient algorithm leverages FPL specifications to achieve up to 500% better sample efficiency compared to Soft Actor Critic. Notably, this work constitutes the first implementation of a non-linear utility scalarization design, intended explicitly for continuous control problems.
Cite this paper
@inproceedings{FPL_IROS25,
title = {{Closing the Intent-to-Reality Gap via Fulfillment Priority Logic}},
author = {El Mabsout, Bassel and Abdelgawad, Abdelrahman and Mancuso, Renato},
booktitle = {IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2025)},
year = 2025,
address = {Hangzhou, China},
url = {https://cs-people.bu.edu/rmancuso/files/papers/FPL_IROS25.pdf}
}
TY - CONF AU - El Mabsout, Bassel AU - Abdelgawad, Abdelrahman AU - Mancuso, Renato TI - Closing the Intent-to-Reality Gap via Fulfillment Priority Logic T2 - IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2025) PY - 2025 CY - Hangzhou, China AB - Practitioners designing reinforcement learning policies face a fundamental challenge: translating intended behavioral objectives into representative reward functions. This challenge stems from behavioral intent requiring simultaneous achievement of multiple competing objectives, typically addressed through labor-intensive linear reward composition that yields brittle results. Consider the ubiquitous robotics scenario where performance maximization directly conflicts with energy conservation. Such competitive dynamics are resistant to simple linear reward combinations. In this paper, we present the concept of objective fulfillment upon which we build Fulfillment Priority Logic (FPL). FPL allows practitioners to define logical formulae representing their intentions and priorities within multi-objective reinforcement learning. Our novel Balanced Policy Gradient algorithm leverages FPL specifications to achieve up to 500% better sample efficiency compared to Soft Actor Critic. Notably, this work constitutes the first implementation of a non-linear utility scalarization design, intended explicitly for continuous control problems. ER -
%0 Conference Paper %A El Mabsout, Bassel %A Abdelgawad, Abdelrahman %A Mancuso, Renato %T Closing the Intent-to-Reality Gap via Fulfillment Priority Logic %B IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2025) %D 2025 %C Hangzhou, China %X Practitioners designing reinforcement learning policies face a fundamental challenge: translating intended behavioral objectives into representative reward functions. This challenge stems from behavioral intent requiring simultaneous achievement of multiple competing objectives, typically addressed through labor-intensive linear reward composition that yields brittle results. Consider the ubiquitous robotics scenario where performance maximization directly conflicts with energy conservation. Such competitive dynamics are resistant to simple linear reward combinations. In this paper, we present the concept of objective fulfillment upon which we build Fulfillment Priority Logic (FPL). FPL allows practitioners to define logical formulae representing their intentions and priorities within multi-objective reinforcement learning. Our novel Balanced Policy Gradient algorithm leverages FPL specifications to achieve up to 500% better sample efficiency compared to Soft Actor Critic. Notably, this work constitutes the first implementation of a non-linear utility scalarization design, intended explicitly for continuous control problems.
[
{
"id": "FPL_IROS25",
"type": "paper-conference",
"title": "Closing the Intent-to-Reality Gap via Fulfillment Priority Logic",
"author": [
{
"family": "El Mabsout",
"given": "Bassel"
},
{
"family": "Abdelgawad",
"given": "Abdelrahman"
},
{
"family": "Mancuso",
"given": "Renato"
}
],
"container-title": "IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2025)",
"issued": {
"date-parts": [
[
2025
]
]
},
"publisher-place": "Hangzhou, China",
"abstract": "Practitioners designing reinforcement learning policies face a fundamental challenge: translating intended behavioral objectives into representative reward functions. This challenge stems from behavioral intent requiring simultaneous achievement of multiple competing objectives, typically addressed through labor-intensive linear reward composition that yields brittle results. Consider the ubiquitous robotics scenario where performance maximization directly conflicts with energy conservation. Such competitive dynamics are resistant to simple linear reward combinations. In this paper, we present the concept of objective fulfillment upon which we build Fulfillment Priority Logic (FPL). FPL allows practitioners to define logical formulae representing their intentions and priorities within multi-objective reinforcement learning. Our novel Balanced Policy Gradient algorithm leverages FPL specifications to achieve up to 500% better sample efficiency compared to Soft Actor Critic. Notably, this work constitutes the first implementation of a non-linear utility scalarization design, intended explicitly for continuous control problems."
}
]
B. El Mabsout, A. Abdelgawad, and R. Mancuso, “Closing the Intent-to-Reality Gap via Fulfillment Priority Logic,” in IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2025), 2025.
El Mabsout, B., Abdelgawad, A., & Mancuso, R. (2025). Closing the Intent-to-Reality Gap via Fulfillment Priority Logic. In IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2025).