Modular autonomous driving systems must coordinate global progress objectives with local safety-driven reactions under imperfect sensing and strict real-time constraints. This paper presents a ROS2-native arbitration module that continuously fuses the outputs of two unchanged and interpretable controllers: a global reference-tracking controller based on Pure Pursuit and a reactive LiDAR-based Gap Follow controller. At each control step, both controllers propose Ackermann commands, and a PPO-trained policy predicts a continuous gate from a compact feature observation to produce a single fused drive command, augmented with practical safety checks. For comparison under identical ROS topic inputs and control rate, we implement a lightweight sampling-based predictive baseline. Robustness is evaluated using a ROS2 impairment protocol that injects LiDAR noise, delay, and dropout, and additionally sweeps forward-cone false short-range outliers. In a repeatable close-proximity passing scenario, we report safe success and failure rates together with per-step end-to-end controller runtime as sensing stress increases. The study is intended as a command-level robustness evaluation in a modular ROS2 setting, not as a replacement for planning-level interaction reasoning.
TABLE I: ROS2 stack and topic interfaces used in simulation experiments.
Fig. 1: System overview. Pure Pursuit (PP) provides global reference tracking, while Gap Follow (GF) provides local LiDAR-based reactive collision avoidance. A PPO-trained arbiter outputs a continuous gate α∈[0,1] that fuses their candidate Ackermann commands into a single drive command. A safety monitor can override the fused command when clearance or data-health checks fail.
Fig. 2: Qualitative interaction sequence in simulation. The ego vehicle transitions from global reference tracking to a local free-space maneuver to pass the leading vehicle, then returns to reference tracking.
Method
SR (%)
Coll. (%)
OffTrk (%)
T/O (%)
Mean runtime (ms)
Worst runtime (ms)
Arbiter (PP+GF)
100.0
0.0
0.0
0.0
31.95
47.12
Sampling MPC
90.0
0.0
0.0
10.0
24.15
61.83
TABLE V: Baseline interaction performance on IMS in F1TENTH Gym ROS ( Nheats=10 ).
Fig. 3: Robustness under base LiDAR impairments (noise, 200 ms delay, 0.3 dropout) with front-cone outliers swept over pout∈{0,0.2,0.4} .
Autonomous driving develops vehicles that perceive road scenes, predict interactions, and plan and control safe movement with limited or no human input.
Current end-to-end autonomous driving systems predominantly rely on frame-based sensors, which suffer from inherent perception latency and motion blur during highly dynamic encounters, specifically sudden pedestrian crossings. To address this critical safety vulnerability, we propose DeepIPCv3, a novel multi-modal autonomous navigation framework that synergizes the dense 3D spatial geometry of LiDAR point clouds with the microsecond-level asynchronous event streams of a Dynamic Vision Sensor (DVS). We introduce a Transformer-inspired cross-modal attention mechanism to dynamically correlate these distinct modalities, allowing the network to instantaneously prioritize high-speed dynamic updates without sacrificing structural scene awareness. The fused latent representations are then mapped to safe local waypoints and executable control commands via a hybrid policy network that blends heuristic trajectory tracking with direct neural predictions. Due to the severe physical risks associated with live testing of these sudden crossing scenarios, the framework is rigorously evaluated offline using a custom multi-modal dataset collected across both well-illuminated noon and challenging evening conditions. Extensive comparative and ablation studies demonstrate that DeepIPCv3 achieves state-of-the-art predictive performance. By effectively eliminating exposure failures and motion blur, the proposed LiDAR and DVS fusion yields the lowest trajectory and control command errors, enabling highly reactive, mathematically bounded evasive maneuvers regardless of ambient illumination. To support future research, we will release the codes to our GitHub repo at https://github.com/oskarnatan/DeepIPCv3.
Oskar Natan, Andi Dharmawan, Aufaclav Zatu Kusuma Frisky +2
Department of Computer Science and Electronics, Universitas Gadjah Mada, Yogyakarta 55281, Indonesia
This paper introduces a novel modular architecture for ROS2 that decouples the logic required to acquire, validate, and interpolate references from the control laws that track them. The design includes a dedicated component, named Reference Generator, that receives references, in the form of either single points or trajectories, from external nodes (e.g., planners), and writes single-point references at the controller's sampling period via the existing ros2_control chaining mechanism to downstream controllers. This separation removes duplicated reference-handling code from controllers and improves reusability across robot platforms. We implement two reference generators: one for handling joint-space references and one for Cartesian references, along with a set of new controllers (PD with gravity compensation, Cartesian pose, and admittance controllers) and validate the approach on simulated and real Universal Robots and Franka Emika manipulators. Results show that (i) references are tracked reliably in all tested scenarios, (ii) reference generators reduce duplicated reference-handling code across chained controllers to favor the construction and reuse of complex controller pipelines, and (iii) controller implementations remain focused only on control laws.
Davide Risi, Vincenzo Petrone, Antonio Langella +3
Department of Information Engineering, Electrical Engineering and Applied Mathematics (DIEM), University of Salerno, 84084 Fisciano, Italy
iG-LIO is a tightly-coupled LiDAR-inertial odometry system fusing generalized-ICP and point-to-plane constraints in an iterated error-state Kalman filter over an incremental voxel map. We report an open-source ROS 2 Jazzy port of the original ROS 1 implementation and, more importantly, the diagnosis of environment-induced numerical failures that appear only after the port: a mechanically faithful migration -- estimation mathematics left unchanged -- compiled and ran, yet diverged with NaN internal values. Both causes trace to the modern ROS 2 toolchain, not the algorithm: a Quality-of-Service (QoS) mismatch that silently drops and reorders IMU samples, and an uninitialized parallel-reduce accumulator arising from the oneTBB + Eigen combination shipped with current distributions. We further correct Ouster point-field parsing to ensure correct point cloud undistortion with newer Ouster revisions, add Velodyne Velarray M1600 support, provide both a compile-time-gated Livox CustomMsg path and a driver-free path for Livox sensors publishing standard PointCloud2 (e.g. Mid-360), and expose the runtime via YAML. The result has been validated in an Ouster OS0 Rev7, an Ouster OS1 Rev 7, and a Livox MID-360. This report is a citable reference for the port itself, not a claim on the underlying algorithm [1]. The ROS 2 port of iG-LIO described in this document can be found at https://github.com/Forestry-Robotics-UC/ig_lio/tree/ros2-jazzy.
Afonso E. Carvalho, David Portugal, Paulo Peixoto
Robotics Institute (RI), Carnegie Mellon University (CMU), Pittsburgh, USA. · Institute of Systems and Robotics (ISR), University of Coimbra (UC), Coimbra, Portugal.