Topic
scheduling algorithm
DynAMO: Dynamic Asset Management Orchestration via Topological Multi-Agent Scheduling
A new research paper introduces DynAMO, a deployment-ready engine for LLM-powered agent orchestration in Industry 4.0. Using Plan-then-Execute architecture with sequential and parallel workflows, it achieves 1.6x median latency reduction, rising to 1.8x on parallelizable tasks. Structured context pruning cuts inference time 30%. Tests on AssetOpsBench show robust performance under fault injection.
New Reinforcement Learning Framework WeCAN Improves DAG Scheduling Efficiency in Heterogeneous Environments
Researchers propose WeCAN, an end-to-end reinforcement learning framework for scheduling directed acyclic graphs (DAGs) on heterogeneous resources. The method uses a two-stage single-pass design with a weighted cross-attention encoder and a skip-extended generation map to close optimality gaps. Experiments on TPC-H query DAGs and ML-compiler graphs show improved makespan with inference time comparable to classical heuristics.