Deploying Vision-Language-Action (VLA) models in real-time autonomous driving faces two barriers: the difficulty of safety assurance for black-box end-to-end outputs, and inference tail latency far exceeding the 20 ms control deadline at 50 Hz, which causes stale decisions to produce dangerous actions. The 50 Hz Fast Reflex layer performs lane detection via INT8-quantized YOLOv8-seg and drives the vehicle using a spline-based path planner. The 1-2 Hz Slow Brain (Qwen3-VL-8B-Instruct, LoRA fine-tuned) runs asynchronously and provides semantic decisions to assist driving in long-tail edge cases. The VLA emits only a Bounded Intermediate Representation comprising a lateral offset, speed scale, validity window, confidence score, and an observation frame timestamp. LASA (Latency-Aware Safety Arbitration) arbitrates this representation at 50 Hz, evaluating staleness against the observation timestamp rather than the IR arrival time, together with asymmetric confidence gating and slew-rate limiting. We validate LASA through a decision-level gate ablation and timing measurements on a physical 1/5-scale platform. Per-gate ablation on a synthetic event stream shows that observation-time arbitration contracts the published command envelope under lane ambiguity - lateral-command RMS drops from 23.2 to 5.4 px and mean speed scale from 0.68 to 0.39 -yielding a markedly smoother and more conservative trajectory than an ungated VLA. On hardware, the off-path distractor false-positive stop rate is 0%, and the 50 Hz reflex loop sustains a sub-millisecond serial-write P99 (0.77 ms) with zero clockmonotonicity violations across the pipeline.
Y. Suh, H. Park, H. Park et al.· International Conference on...· 0 citations
Accurate depth perception is a cornerstone of autonomous driving, yet LiDAR sensors—the primary source of metric depth—remain costly and operationally complex. In this paper, we propose a cross-modal supervision framework that uses sparse LiDAR depth maps solely during training, enabling camera-only dense depth inference at test time. A ConvNeXt-base encoder with an FPN neck and a lightweight depth decoding head is trained with a log-scale L1 loss, gradient consistency term applied exclusively at valid LiDAR pixels (~0.7% pixel density), and an image-guided edge-aware smoothness loss operating on all pixels, alongside a two-phase backbone freeze-then-unfreeze strategy to stabilize early convergence. Evaluated on a large-scale Korean highway dataset of 64,840 frames, our model achieves AbsRel of 0.0675, RMSE of 3.907 m, and $\delta \lt 1.25$ accuracy of 0.943, demonstrating that ultra-sparse LiDAR supervision is sufficient to train competitive monocular depth estimators.
H. Park, Young Hoon Suh, Sun-Kug Jung et al.· International Conference on...· 0 citations
Time-Sensitive Networking (TSN) is a key mechanism for deterministic in-vehicle Ethernet, but its end-to-end behavior over 10BASE-T1S shared media remains unclear under mixed traffic in zonal architectures. This paper experimentally evaluates periodic control traffic delivered from an upper zonal Ethernet TSN domain to a lower-tier 10BASE-T1S bus. The testbed connects two cascaded SJA1110-based switches and an EVB-LAN9383 switch to a 10BASE-T1S motor-control node, with best-effort (BE) background traffic generated by an internal processing core of the second SJA1110. Control latency is measured by matching a 32-bit sequence field between source-side Ethernet and destination-side 10BASE-T1S observation points while varying BE load from 0% to 95%. We compare No GB, which strictly separates scheduled traffic (ST) and BE traffic without a guard band; fixed guard-band (GB), which closes the cycle end before the next ST window; Shared, which opens the remaining non-ST interval to both ST and BE traffic; and two ST-slot/GB-width variants. The results show that No GB suffers severe latency growth, whereas fixed GB stabilizes lowand medium-load latency near 3 ms but loses matched delivery under high load. Shared provides the best control-delivery tradeoff in the evaluated testbed, maintaining above 95% matched ratio through $\mathbf{9 0}$ % load with $\mathbf{5 1 0. 8} \mu \mathbf{s}$ and $\mathbf{6 7 8. 3} \mu \mathbf{s}$ mean latency at 0% and 90%, respectively. These results indicate that flexible shared scheduling can improve control-flow delivery compared with strict static partitioning in the evaluated 10BASE-T1S-based zonal topology, while BE fairness and synchronization-related phase effects require further study.
Jae Hun Kang, Y. Do, C. Park et al.· International Conference on...· 0 citations
