Multi-Agent Scheduling of Instrument-Time for High-Throughput Palladium-Catalysed Cross-Coupling
Abstract
NMR time and a shared liquid-handling robot are scarce, noisy and only partially observable, so the practical question is which experiment to run next on which instrument. Existing reaction-optimisation benchmarks measure budgets in experiments rather than instrumenthours, and they learn chemical feasibility only by paying for infeasible runs. We present CHASE, a centralised-training decentralised-execution scheduler with three coupled modules: a belief encoder (PEB) infers per-instrument throughput and channel health from local event histories and exchanges only low-bandwidth summaries; a chemistry feasibility gate (CFG) masks chemically impossible actions with a four-layer compatibility graph; and a counterfactual attribution module (CFA) converts delayed campaign reward into per-step credit through leave-one-experiment-out surrogate scoring. On the Buchwald–Hartwig (3,955 reactions) and Suzuki–Miyaura (5,760 nanomole-scale conditions) HTE datasets, the ten Olympus benchmarks and a plant-wide Tennessee Eastman process, CHASE reaches 94.1% best-found yield under a shared 60 instrument-hour budget, +3.3 points over the strongest same-setting baseline, attains 90% of the attainable optimum with 25.6% fewer experiments, and cuts infeasible actions to 0.7–1.6% across plate, flow and fault-injection settings.
Related
- Nickel-catalysed synthesis of symmetrical and unsymmetrical 2,2′-bipyridines via cross-electrophile coupling (XEC) using manganese reductant
- Auto‐Click Synthesis of a Copper(I) –N ‐Heterocyclic Carbene Porous Organic Polymer and Its Transmetalation to Palladium(II) for Azide–Alkyne Cycloaddition and Cross‐Coupling Reactions
- Room temperature Suzuki coupling of aryl iodides, bromides, and chlorides using a heterogeneous carbon nanotube-palladium nanohybrid catalyst
- Synthesis of 5-[3-amino-imidazo[1,2- a ]pyridin-6-yl]pyridin-2-ones and derivatives through the combination of palladium-catalyzed cross-coupling and GBB multi-component reactions
- Galvanic Coupling Links Heterogeneous and Molecular Palladium Catalysis in Aerobic Alcohol Oxidation
BibTeX
@misc{shao2026multiagent,
title = {{Multi-Agent Scheduling of Instrument-Time for High-Throughput Palladium-Catalysed Cross-Coupling}},
author = {Yuhang Shao and Siqi Meng and Yutian Luo and Jiawei Tan and Chengyu He},
howpublished = {ChemRxiv},
year = {2026},
month = oct,
doi = {10.26434/chemrxiv.15010837/v1},
url = {https://chemrxiv.org/doi/full/10.26434/chemrxiv.15010837/v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Multi-Agent Scheduling of Instrument-Time for High-Throughput Palladium-Catalysed Cross-Coupling
AU - Yuhang Shao
AU - Siqi Meng
AU - Yutian Luo
AU - Jiawei Tan
AU - Chengyu He
PY - 2026
DA - 2026/10/22
PB - ChemRxiv
DO - 10.26434/chemrxiv.15010837/v1
UR - https://chemrxiv.org/doi/full/10.26434/chemrxiv.15010837/v1
AB - NMR time and a shared liquid-handling robot are scarce, noisy and only partially observable, so the practical question is which experiment to run next on which instrument. Existing reaction-optimisation benchmarks measure budgets in experiments rather than instrumenthours, and they learn chemical feasibility only by paying for infeasible runs. We present CHASE, a centralised-training decentralised-execution scheduler with three coupled modules: a belief encoder (PEB) infers per-instrument throughput and channel health from local event histories and exchanges only low-bandwidth summaries; a chemistry feasibility gate (CFG) masks chemically impossible actions with a four-layer compatibility graph; and a counterfactual attribution module (CFA) converts delayed campaign reward into per-step credit through leave-one-experiment-out surrogate scoring. On the Buchwald–Hartwig (3,955 reactions) and Suzuki–Miyaura (5,760 nanomole-scale conditions) HTE datasets, the ten Olympus benchmarks and a plant-wide Tennessee Eastman process, CHASE reaches 94.1% best-found yield under a shared 60 instrument-hour budget, +3.3 points over the strongest same-setting baseline, attains 90% of the attainable optimum with 25.6% fewer experiments, and cuts infeasible actions to 0.7–1.6% across plate, flow and fault-injection settings.
ER -