analiza aiqbench
This paper treats neural-network inference as a real-time scheduling problem for robots with tight compute budgets. Its main contribution is a way to split end-to-end deadlines into per-node sub-deadlines, with explicit conditions under which those guarantees hold. It also separates run-level completion from per-node deadline success, a distinction that matters for anyone judging whether an edge AI stack is safe to deploy. The evidence is limited to three Orin and M3 Max platforms, and the timing claims depend on profiled bounds that may not hold outside the tested scenarios.
MERIS: Adaptive Real-Time Scheduling of Refined Inference DAGs in Dynamic Robotic Environments
arXiv:2605.24044v2 Announce Type: replace Abstract: Robots deployed in dynamic environments execute changing directed-acyclic-graph (DAG) workloads under limited compute and timing budgets. This article presents MERIS, a framework for model scheduling under real-time constraints that combines proportional intermediate deadlines, MIMONet-aware graph refinement, and event-driven synchronization for multi-task inference on resource-constrained robotic platforms. We distinguish three representations: the application DAG, each MIMONet's internal encoder/decoder graph, and the refined scheduling DAG, and make the scope of the timing argument explicit. If the profiled execution bounds, release assumptions, precedence barriers, and per-level capacity premises hold, satisfaction of the assigned sub-deadlines composes to the end-to-end deadline. For non-partitionable tasks, we define a profile-bounded admission rule; before interpreting that rule as empirical timing evidence, we require direct per-task completion checks. We report run-level QoE alongside the offered-node deadline missing rate, making explicit that run completion and successful service of every node are different outcomes. We propose a comprehensive minibenchmark for evaluating the performance of schedulers under various DNN-driven autonomous driving scenarios. Extensive experiments on Nano Orin, AGX Orin, and M3 Max demonstrate the effectiveness of our proposed method.
Abstract:Robots deployed in dynamic environments execute changing directed-acyclic-graph (DAG) workloads under limited compute and timing budgets. This article presents MERIS, a framework for model scheduling under real-time constraints that combines proportional intermediate deadlines, MIMONet-aware graph refinement, and event-driven synchronization for multi-task inference on resource-constrained robotic platforms. We distinguish three representations: the application DAG, each MIMONet's internal encoder/decoder graph, and the refined scheduling DAG, and make the scope of the timing argument explicit. If the profiled execution bounds, release assumptions, precedence barriers, and per-level capacity premises hold, satisfaction of the assigned sub-deadlines composes to the end-to-end deadline. For non-partitionable tasks, we define a profile-bounded admission rule; before interpreting that rule as empirical timing evidence, we require direct per-task completion checks. We report run-level QoE alongside the offered-node deadline missing rate, making explicit that run completion and successful service of every node are different outcomes. We propose a comprehensive minibenchmark for evaluating the performance of schedulers under various DNN-driven autonomous driving scenarios. Extensive experiments on Nano Orin, AGX Orin, and M3 Max demonstrate the effectiveness of our proposed method.
Robotics (cs.RO); Software Engineering (cs.SE); Systems and Control (eess.SY)
arXiv:2605.24044 [cs.RO]
(or arXiv:2605.24044v2 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2605.24044 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 275 cuvinte
Subiecte conexe
EnvSched: Environment-Aware Scheduling for Rollout-Efficient Agent Reinforcement Learning
arXiv · 28 oct. 2026
Quantum enhances the HPC stack
Pasqal · 22 iun. 2026
Environment Evolution for Terminal Agents
arXiv · 5 sept. 2026
A Learning Method with Gap-Aware Generation for Heterogeneous DAG Scheduling
arXiv · 16 iun. 2026