Wodtko, Thomas (Ulm University), Scheible, Alexander (Ulm University), Authaler, Dominik (Ulm University), Buchholz, Michael (Universität Ulm)

Batched Minimal Latency In-Sequence Ordering for Multi-Channel Data

Scheduled for presentation during the Regular Session "S28a-Multi-Sensor Fusion and Perception for Robust Autonomous Driving" (TH-LM-T28), Thursday, November 20, 2025, 10:30−10:50, Stradbroke

2025 IEEE 28th International Conference on Intelligent Transportation Systems (ITSC), November 18-21, 2025, Gold Coast, Australia

This information is tentative and subject to change. Compiled on October 18, 2025

Abstract

Fusion is a crucial part of the data processing in autonomous vehicles. Data from multiple sensors must be combined reliably to allow resilient behavior and motion planning. Generally, sensor data is received in an arbitrary order due to different acquisition frequencies, followed by transmission delays and varying processing times. However, processing and fusion modules usually require in-sequence data reception and may impose further requirements, such as receiving data batches containing data from different sensors acquired at similar times. While approaches exist to ensure in-sequence data reception, methods are lacking to realize any other data reception requirements. With this work, we propose a novel buffering approach that extends our previously proposed minimal latency method. Our new approach operates in two modes of delivering groups of data samples, providing wide applicability for various tasks. Always considering the estimated data acquisition sequence, data from streams is either accumulated and forwarded in a batch or matched, and tuples with one data sample from every stream are forwarded. Our method explicitly detects missing data samples, is robust to unreliable transmission channels, and can even compensate for missing data in some cases while maintaining minimal induced latency. In our evaluation, we demonstrate the superiority of our approach over other state-of-the-art methods and highlight the improvements toward robust fusion in intelligent vehicles.