Calibration-First Reward-Component Auditing for Reinforcement Learning Control in Smart Greenhouses
Quick Answer
The proposed calibration-first reward audit framework enhances greenhouse reinforcement learning by decomposing scalar rewards into actionable components like temperature and humidity.
Quick Take
This framework, implemented in GreenLight-Gym, allows for consistent evaluation across various training and deployment scenarios, improving control strategies in smart greenhouses.
Key Points
- Introduces a reproducible framework for auditing greenhouse control rewards.
- Decomposes rewards into components like temperature, CO2, and humidity.
- Adapts GreenLight to the second Autonomous Greenhouse Challenge climate traces.
- Scores greenhouse control components using logged data for better insights.
- Facilitates faster testing of climate-control ideas in smart greenhouses.
Paper Resources
📖 Reader Mode
~2 min readAbstract:Greenhouse reinforcement learning can test climate-control ideas at a speed and scale that is difficult to achieve with crop experiments alone. For smart-greenhouse control, however, a single simulator return is not enough: a grower or control engineer also needs to know when the policy heats, enriches CO2, vents, manages humidity, deploys screens, or uses this http URL propose a reproducible calibration-first reward audit framework that keeps named greenhouse-control reward components comparable across simulator training, facility-adapted rollouts, logged Autonomous Greenhouse Challenge records, and actuator-rule distillation. In GreenLight-Gym, the framework decomposes the scalar reward into conditional temperature, CO2, humidity and vapor-pressure-deficit, screen, and actuation-proxy terms; adapts GreenLight to the second Autonomous Greenhouse Challenge logged climate traces; and scores the same components on logged greenhouse data.
| Comments: | 28 pages, 8 figures |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2607.11959 [cs.AI] |
| (or arXiv:2607.11959v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2607.11959 arXiv-issued DOI via DataCite |
Submission history
From: Yaojun Wang [view email]
[v1]
Sun, 12 Jul 2026 07:37:54 UTC (528 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.AI
See more →HOBA: Hierarchical On-Policy Bidding Agents for Adaptive Online Advertising
HOBA (Hierarchical On-policy Bidding Agents) is a novel hierarchical reinforcement learning framework that enhances online advertising bidding systems by improving adaptability and reducing hyperparameter tuning costs. It utilizes a for hyperparameter inference, a SARSA agent for expert model selection, and a dynamic expert pool for bid execution, achieving a +3.6% increase in target cost during large-scale deployment and outperforming state-of-the-art baselines on AuctionNet.