EasyNav: A Modular Architecture for Real-Time Navigation Across Heterogeneous Robot Environments
Authors: Francisco Martín Rico, Francisco Miguel Moreno, Juan Sebastián Cely Gutiérrez, Esther Aguado, Miguel Ángel De Miguel Paraíso, José Miguel Guerrero
Organizations: Intelligent Robotics Lab, Universidad Rey Juan Carlos, Fuenlabrada, Madrid, Spain
Most robot navigation stacks implicitly bind mapping, localization, planning, and control to specific representations and data flows, yielding architectures hard to extend, compare experimentally, or reason about end-to-end timing. This is critical in heterogeneous scenarios, where multiple spatial representations, sensing modalities, and coordinate frames coexist and evolve at different rates. This paper presents a navigation architecture centered on an explicit, shared navigation state that serves as the sole interaction interface between modular components. Constraining mapping, localization, planning, and control to operate over this shared state enables interchangeable, representation-agnostic pipelines under explicit compatibility and timing contracts. We realize this architecture as EasyNav, an open source ROS 2 framework, and validate it through real-robot experiments, including indoor navigation with laser sensing and large-scale indoor-outdoor navigation with GPS and 3D LiDAR across multi-level routes. Compared with Nav2 under matched conditions, EasyNav reduces reaction latency from 377 to 33 ms indoors and from 71 to 8 ms outdoors, while also using less CPU and memory.
Figures & tables
Fig. 1 : Examples of navigation scenarios enabled by the proposed architecture.
Property
Fawkes
Nav2
Orocos
CLARAty
EasyNav
[ 8 ]
[ 1 , 9 ]
[ 13 ]
[ 5 ]
Global nav-state R/W contract
partial
✗
n/a
n/a
✓
Per-entry stamp/frame
partial
✗
n/a
n/a
✓
Heterog. repr., 1 pipeline
n/a
partial
n/a
n/a
✓
Perception-triggered exec.
✗
✗
partial
n/a
✓
Explicit RT/non-RT split
✗
✗
✓
partial
✓
TABLE I : Comparison with representative related systems. ✓: supported; ✗: not supported; partial : supported in a restricted sense (see Section II ); n/a: not applicable to that system’s scope.
Fig. 2 : Functional roles interact exclusively through the shared, time-stamped navigation state; no direct dependencies exist between roles.
Fig. 3 : EasyNav Design. The NavState blackboard (lavender box) stores the information that plugins read and write. Sensors is the single entry point through which sensory data enters the system. Colored puzzle pieces represent the plugin placeholders of each functional role, which must fit together to form a valid pipeline. Navigation goals reach the system through the GoalManager , which writes them into NavState for the Planner to consume. Red boxes indicate execution frequencies in the real-time cycle, and green boxes indicate execution frequencies in the non-real-time cycle.
Indoor
Outdoor
Platform
TurtleBot 4
Summit XL
Hardware
Raspberry Pi 4
i7-10700
Middleware
Fast DDS
Fast DDS + Zenoh bridge
Sensing
LiDAR 2D
LiDAR 3D + IMU + GNSS
Env. repr.
Costmap
Costmap
Localization input
Range + odom.
IMU + odom. + GNSS
TABLE II : Experimental platforms and navigation pipelines.
Fig. 4 : Comparison of CPU utilization and latency on both experiments. Outdoor CPU boxes summarize within-trial samples from a single continuous run, not independent executions (Section V-B ).
Scenario
Metric
EasyNav
Nav2
Diff.
Indoor
CPU [%]
24.38±1.35
36.01±1.81
−11.63 pp
Memory [MB]
115.83±0.60
189.96±0.65
−74.13
Cmd freq. [Hz]
49.41±0.71
18.11±0.87
+31.30
Cmd ud. freq. [Hz]
23.02±3.03
16.23±1.92
+6.79
Travelled dist. [m]
142.23±0.67
132.09±0.44
+10.14
Latency [ms]
32.97±7.20
377.19±80.16
−344.21
TABLE III : Summary of the experimental results. Values are reported as mean ± standard deviation across independent runs, except for outdoor CPU/memory/frequency rows, where they describe sample-level variability within a single continuous trial (Section V-B ).
The deployment of mobile robots in large-scale, multi-floor environments demands navigation systems that achieve spatial scalability without compromising local kinematic precision. Traditional navigation stacks, reliant on monolithic occupancy grid maps, face severe bottlenecks in storage efficiency, cross-floor reasoning, and long-horizon planning. To address these limitations, this paper presents osmAG-Nav, a complete, open-source ROS2 navigation stack built upon the hierarchical semantic topometric OpenStreetMap Area Graph (osmAG) map standard. The system follows a "System of Systems" architecture that decouples global topological reasoning from local metric execution. A Hierarchical osmAG planner replaces dense grid searches with an LCA-anchored pipeline on a passage-centric graph whose edge costs derive from local raster traversability rather than Euclidean distance, yielding low-millisecond planning on long campus-scale routes. A Rolling Window mechanism rasterizes a fixed-size local metric grid around the robot, keeping the local costmap memory footprint independent of the total mapped area, while a Segmented Execution strategy dispatches intermediate goals to standard ROS2 controllers for smooth handoffs. System robustness is reinforced by a structure-aware LiDAR localization framework that filters dynamic clutter against permanent architectural priors. Extensive experiments on a real-world multi-story indoor-outdoor campus (>11,025 m^2) show that, on the same-floor benchmark subset, osmAG-Nav delivers up to 7816x lower planning latency than a grid-based baseline on long routes while maintaining low path-length overhead and lifelong localization stability. A single-floor long-range robot mission further validates the integrated stack reliability. The full stack is released as modular ROS2 Lifecycle Nodes.
Yongqi Zhang, Jiajie Zhang, Chengqian Li +2
School of Computer Science and Technology, ShanghaiTech University, 393 Huaxia Road, Shanghai, 201210, Shanghai, China.
Long-term navigation for service robots faces crit- ical challenges like the accumulation of odometry drift and sensor error, which progressively degrade 2D maps and renders traditional path planning algorithms (e.g., A*, RRT*, DiPPer, ViT-A*) ineffective over time. To address this, we propose a user-friendly, interactive framework that eliminates the reliance on globally consistent maps. Our approach integrates visual perception with Large Language Models (LLM) to interpret user commands via text or voice. Instead of relying on a drift- prone global map, the system generates a sequential action plan based on local visual cues and egocentric geometric instructions. These action plans are executed sequentially, allowing the robot to navigate known and unknown environments safely. By reset- ting localization relative to immediate targets, our framework effectively works with a minimum accumulation drift strategy, ensuring accurate, efficient, and collision-free navigation without the maintenance overhead of traditional mapping. Experiments on real-world and simulated data have shown significant improve- ments over other methods. Our source code is publicly accessible at https://github.com/PraveenSingh24/VL-Navigation.
Praveen Kumar, K. R. Guruprasad, Tushar Sandhan
Department of Electrical Engineering, Indian Institute of Technology Kanpur, India · Department of Mechanical Engineering, Indian Institute of Technology Kanpur, India
Lifelong embodied navigation in dynamic environments requires robots to form persistent scene understanding from fragmentary observations, which remains difficult for existing methods that rely on explicit maps or scene graphs and struggle to generalize beyond structured settings. We propose AllDayNav, a lifelong self-learning navigation framework that implicitly encodes scene dynamics into the billion-scale parameters of a large model via reinforcement learning, powered by a self-evolving multimodal memory that maintains and updates visual keyframes, semantic descriptions, and temporal context while autonomously generating open-vocabulary instructions, image goals, and structured rewards. Experiments in both synthetic and real-world environments across cross-room, cross-episode, and cross-task scenarios show that AllDayNav achieves success rates approaching 100% and consistently surpasses strong map-based, VLM, and RL baselines in path efficiency and robustness, demonstrating implicit, memory-driven reinforcement learning as a scalable alternative to explicit mapping for reliable lifelong navigation.
Hang Yin, Yinan Liang, Jiazhao Zhang +4
Tsinghua University, Beijing 100084, China · Galbot Robotics, Beijing, China · Peking University, Beijing 100871, China +1