Papers for

mobile robot developers

Papers whose findings have a practical use for this group, as judged from the abstract. Open a paper to read what it means in practice.

Mobile robots navigate safer using uncertainty-aware depth from one camera

From Depth to Safe Depth: Uncertainty-Aware Monocular Perception for Mobile Robot Navigation

Abstract: Monocular depth foundation models make it tempting to replace depth sensors on mobile robots with a single RGB camera. We show that, used off the shelf, this is unsafe: in a simulated indoor navigation benchmark, a standard Nav2 stack driven by Depth Anything V2 (metric) collides in 63\% of 120 episodes, because the model systematically over-estimates the distance to nearby obstacles. We propose a lightweight plug-in module that turns predicted depth into \emph{safe depth}. A 0.58\,M-parameter head on top of the \emph{frozen} depth model predicts, per pixel, a log-depth correction $μ$ and an uncertainty $σ$, trained with a $β$-weighted Gaussian negative log-likelihood. The navigation stack consumes the one-sided bound $\hat d_{\text{safe}}=d\,e^{μ-kσ}$, where $k$ is set by conformal calibration on a held-out world, and uncertainty also decides which pixels may be treated as floor. The correction halves the depth error (AbsRel $0.262\rightarrow0.134$) and reduces near-range over-estimation from 59.8\% to 12.0\%. In navigation, safe depth raises the success rate from 37\% to 70\% and lowers collisions from 63\% to 3\%, close to the 2\% of ground-truth depth. A constant safety margin calibrated to the same pixel coverage reaches only 30\% success because it freezes the robot; uncertainty-aware safety wins 52 of 56 discordant paired episodes against it. Gains hold on glass, texture-less walls and dark regions.

Tue 10 NovRoboticsComputer Vision and Pattern Recognition
The gist
Monocular depth models that estimate distances using just one RGB camera can make mobile robots think obstacles are farther away than they really are, causing crashes. The authors show this is unsafe if these models are used without adjustment. They introduce a small add-on that corrects depth predictions by estimating both a correction and an uncertainty for each pixel. This correction makes the robot’s navigation much safer and more reliable, almost matching the performance of using real depth sensors.
Open → 2611.13769v1

Single radar captures 3D views both sides for mobile robots

Single-Radar Bilateral 3D Near-Field Imaging for Mobile Robots

Abstract: Mobile robots operating in homes need reliable 3D perception of nearby furniture and surrounding structures, yet recovering this information on both sides of the robot's path with a single forward-mounted millimeter-wave radar remains challenging because of near-field wavefront curvature, viewpoint-dependent target visibility, and limited elevation diversity. This paper presents a motion-assisted framework for reconstructing structures on both sides of a robot's trajectory. Our near-field model relates scene reflectivity to stepped-frequency multiple-input multiple-output (MIMO) observations across frequencies and robot positions using spherical paths from transmitter (Tx) to scatterer to receiver (Rx). Bilateral Tx combining provides wideband directional weighting, while sliding-aperture back-projection coherently focuses the measurements using near-field propagation distances. Top-K power fusion combines local images with different visibility. The reconstruction provides lateral-forward support for multi-layer height extraction, followed by guarded adaptive height refinement using the original MIMO channels. Experiments in controlled, office/study, and corridor environments produce mean lateral-forward localization errors of 4.65, 5.63, and 5.98 cm, respectively, with peak-to-background ratios (PBRs) exceeding 13 dB. The framework recovers bilateral 3D scattering structure from one forward pass using a single radar, without relying on learned representations of the scene.

Sun 8 NovRoboticsComputer Vision and Pattern Recognition
The gist
Mobile robots need to see their surroundings clearly on both sides, but using just one radar facing forward makes this hard. The authors developed a new technique that moves the robot and uses special radar signals to create detailed 3D images of objects on either side. Their method processes radar data carefully to focus on nearby objects and builds accurate maps showing shapes and heights. They tested it in real indoor areas and achieved good precision without relying on previous scene knowledge.
Open → 2611.11296v1

Gpu-powered mapping speeds up robot motion planning indoors and outdoors

GPU-accelerated Incremental Euclidean Distance Transform for Online Motion Planning of Mobile Robots

Abstract: In this letter, we present a volumetric mapping system that effectively calculates Occupancy Grid Maps (OGMs) and Euclidean Distance Transforms (EDTs) with parallel computing. Unlike these mappers for high-precision structural reconstruction, our system incrementally constructs global EDT and outputs high-frequency local distance information for online robot motion planning. The proposed system constructs OGM with massive amount of data from multiple types of sensor inputs. Using GPU programming techniques, the system quickly computes EDT in parallel within local volume. The new observation is continuously integrated into the global EDT using the parallel wavefront algorithm while preserving the historical observations. Experiments with datasets have shown that our proposed approach outperforms existing state-of-the-art robot mapping systems and is particularly suitable for mapping unexplored areas. In its actual implementations on aerial and ground vehicles, the proposed system achieves real-time performance with limited onboard computational resources.

Fri 30 OctRobotics
The gist
Planning safe paths for robots needs up-to-date maps that show obstacles and distances. The authors created a system that quickly builds these maps using many sensors and powerful graphics processors (GPUs). This system updates maps bit by bit as new data arrives, keeping old information too. It helps flying and ground robots move around in unfamiliar places without delays, even with limited computer power onboard.
Open → 2610.41667v1

Compact robot agents match large planners with self improved harnesses

TinyAstral: Transferring Self-Evolving Harnesses to Tiny Generalist Manipulation Agents

Abstract: Robot agents built on vision-language models depend on both the planner and its harness: the tools, feedback, memory, and recovery logic around it. Can improvements to a large teacher's harness be carried into a compact agent with lower inference compute cost? We present TinyAstral, a two-stage approach. We first freeze an open-weight teacher and improve its harness through recursive self-improvement (RSI): frontier LLM agents read the teacher's trajectories and propose harness edits, and each edit is accepted only if it passes a paired statistical test. We then freeze the final harness, collect teacher trajectories with reasoning and failure recovery, and distill a small student that replaces the teacher in the same harness. On RoboTwin, harness evolution raises the teacher's success from 33.4% to 63.6%; the final 4B student reaches 62.8%. Across four harness versions, distilled 4B students score close to their corresponding teachers, and later versions yield stronger students, showing that harness evolution gains carry into compact models. The harness gains also transfer to other frozen planners without retraining. On identical episodes, the final 4B student outperforms open 4B agents, the open 27B harness-RSI model SenseNova-RoboRSI, and our own teacher model placed in three competitors' harnesses. It stays close to its teacher on all three training benchmarks, and on three benchmarks never used for RSI or training it scores close to a 552B planner running the same harness. On the measured deployment episodes, it uses about one fifth of its teacher's accelerator time.

Fri 30 OctRoboticsComputer Vision and Pattern Recognition
The gist
Robot agents that use vision and language models rely heavily on a supportive framework called a harness, which includes tools and strategies for planning and correcting actions. The authors show how improvements made to a large, powerful robot planner’s harness can be transferred to a much smaller, faster robot. They do this by letting language models suggest improvements to the harness, testing those improvements, then training a smaller robot to use the improved harness and mimic the large planner’s behavior. The smaller robot achieves nearly the same success as the bigger one but runs much faster.
Open → 2610.41216v1

Monte Carlo sampling helps robots find position quickly in mazes

AFast Robot Localization Method Based on Monte Carlo Sampling

Abstract: Localization is an essential capability of intelligent robots. However, in real-world constrained environments, robots often cannot rely on external localization systems. Therefore, this paper proposes an online particle-filter localization method based on Monte Carlo sampling. The method is applied to the localization of a mobile robot in a maze environment and requires only four range sensors mounted on the front, rear, left, and right sides of the vehicle, without the need for any heading information. Simulation experiments demonstrate that even in a maze where local regions are highly similar, the vehicle can still achieve relatively accurate localization, thereby providing a fast and effective approach for robot localization in complex environments.

Thu 29 OctRobotics
The gist
Knowing where a robot is helps it move safely and complete tasks. But robots often can’t use GPS or other outside help indoors or in tricky places. The authors came up with a method that uses random sampling to estimate the robot’s location using only four distance sensors on different sides. They tested it in a maze with many similar-looking areas and found the robot could still figure out where it was pretty accurately and quickly.
Open → 2610.39708v1

Vision language navigation improved by remembering routes and landmarks

What Should a Navigator Remember? Spatiotemporal Memory for Zero-Shot Vision-Language Navigation

Abstract: Zero-shot vision-and-language navigation (VLN) requires an agent to use past observations without navigation-specific training. We study what information an 8B-scale navigation system should retain and how it should be organized for subsequent decisions. Our agent combines instruction progress, hierarchically compressed observation history anchored to poses, and accumulated geometry in complementary task, temporal, and spatial memories. An 8B observer summarizes observations, and an 8B reasoner uses the memories to select waypoint intents or request additional views, a geometric controller executes the selected targets. A semantic information perspective motivates the distinction between retaining evidence, making it accessible to the reasoner, and acquiring further observations. On R2R-CE val-unseen, the complete system achieves the highest among the evaluated Qwen3-based zero-shot baselines. Removing temporal memory reduces SR and sharply increases average path length from 16.9 to 39.1m, removing the spatial-map and planning component reduces SR as well. Real-world experiments on a Unitree Go2 show that our zero-shot method achieves higher success rates than the post-trained baseline in both corridor and office environments, suggesting better transfer to the evaluated real-world settings. As a complementary evaluation, LAW2 uses paired instruction interventions to examine behavioral responses, instruction following ability beyond endpoint success. These results support preserving task progress, observation history, and geometry in structured memory for zero-shot navigation with smaller open models.

Wed 28 OctRoboticsArtificial Intelligence
The gist
Vision-language navigation means following spoken or written instructions to move through places using cameras and language understanding. This paper studied what an AI system should remember while navigating without special training for the task. The authors used three types of memory: the progress in the task, compressed past views linked to locations, and geometric maps. Their system makes better choices by combining these memories and robotic controls. They showed it works well in both simulated and real-world environments, even beating some trained baselines.
Open → 2610.38318v1

Insect-inspired neural network classifies gases using low power

Online gas classification at the edge

Abstract: Field applications in chemical sensing, from portable safety monitors to mobile robots, require models that are lightweight, fast to respond, and can be recalibrated from limited labelled data. Current gas classification pipelines assume controlled sampling cycles and offline feature extraction, and rely on computationally expensive models. With a few thousand neurons and sub-milliwatt power consumption, insects reliably classify odours in open environments. Inspired by the insect olfactory system, we present a neural network architecture that classifies raw sensor array readings using a sparse, high-dimensional binary representation. We evaluate it on random sequences of single gases and binary mixtures, introducing new samples dynamically and classifying gas identity from the transient sensor response. In this setting, the architecture out-performs a linear \gls{svm} on gas mixtures and matches an optimised non-linear \gls{svm}, while maintaining online class-level frequency estimates that distinguish familiar from novel stimuli. It relies exclusively on binary weights and Hebbian-based local learning, remaining computationally lightweight and well suited to ultra-low-power neuromorphic processors. We demonstrated the network on-chip with a hazardous gas detection experiment running on the mixed-signal neuromorphic chip TEXEL. The on-chip classifier yielded an accuracy of 74.6\% while running at 482 $\unit{\nano\joule}$ per inference and 27 \unit{\micro\watt} of static-power.

Tue 27 OctEmerging TechnologiesMachine Learning
The gist
Detecting different gases quickly and with low power is important for safety and robots in the field. The authors created a tiny neural network inspired by how insects smell, which uses a simple way to represent data and can learn on the spot. It can tell apart single gases and mixed gases from raw sensor data, performing as well as complex existing methods but much more efficiently. They even tested it on a special low-power chip that runs the model with very little energy.
Open → 2610.36999v1

LiDAR based robot navigation improves safety and smoothness near workers

Learning Interaction-Aware Path Corrections from Unlabeled Trajectories: A LiDAR-Based Sim-to-Real GNN Framework for Construction Robots

Abstract: Construction robots operating near workers require navigation that can anticipate human motion while remaining robust to dynamic and unstructured site conditions. However, existing social navigation methods depend on hand-crafted interaction rules, labeled demonstrations, or robot-specific control policies, limiting their scalability and transferability in construction environments. We propose a real-time, LiDAR-only framework that learns interaction-aware path corrections from unlabeled demonstrations. Constant-velocity extrapolation generates residual learning targets, while a GNN models relationships among the robot, pedestrians, and goal context. During deployment, an extended Kalman filter tracks LiDAR detections to construct dynamic graphs, and the GNN offsets planner-generated waypoints through a planner-agnostic interface. Within the bricklaying trial, the GNN reduced path irregularity by 39.2\% and increased minimum TTC. Across Dynablox sequences, it increased minimum TTC by 38.3\% while producing shorter, smoother paths than the baselines. Ablation studies supported a 1.0-s prediction interval and confirmed the benefits of target context, reflection augmentation, and feature normalization.

Sun 25 OctRobotics
The gist
Construction robots need to move safely around people on busy and changing job sites. Current methods often rely on preset rules or require labeled examples, which limits how well they work in new situations. The authors created a way for robots to use only LiDAR sensors to learn how to adjust their paths by watching unlabeled movements of people and robots. Their method uses a graph neural network to understand interactions and improves the robot’s path smoothness and safety in tests.
Open → 2610.34060v1

Lightweight model predicts short-term human paths for mobile robots

Real-Time Human Trajectory Prediction from RGB-D for Mobile Robots Using Bi-Mamba

Abstract: Predicting the short-term future trajectory of a person from RGB-D cameras is a key capability for mobile robots navigating in dynamic and human populated environment. The key challenge to deploying learned predictors on real robots is the train-deploy pipeline gap. Training data are often prepared with an idealized, smoothed ground truth, whereas at inference, the model is fed noisy estimates from pose detectors and depth sensors. The model is therefore evaluated on an input distribution it never encountered during training. Most existing models rely on recurrent neural networks (RNNs) or transformers, which either struggle to capture short-term temporal dependencies or suffer from computational cost. We present TrajMamba, a lightweight, bidirectional state space model for short-term human trajectory prediction using an RGB-D camera. The model uses a single shared processing path for both training and deployment, aligning human pose extraction, depth back projection, and smoothing processes. It encodes a short window of persons's hip center velocities with stacked bidirectional Mamba blocks and decodes future velocities with a GRU head. On the ETH/UCY benchmark, TrajMamba reaches an average ADE of 0.29 m FDE of 0.51 m. With 262K parameters, it runs in real time on modest hardware and is deployed on a TIAGo robot with an onboard NVIDIA Jetson Orin.

Sat 24 OctRoboticsMachine Learning
The gist
Predicting where a person will walk next helps robots move safely around people. The authors created a lightweight computer model called TrajMamba that uses camera images and depth info to guess short-term paths a person might take. Unlike past methods, their model works well even when the input data is noisy and runs fast on small robot computers. They tested it on popular benchmarks and real robots with good accuracy and speed.
Open → 2610.33916v1

Memory aware management keeps dense maps on small robots working

Memory-Aware Keyframe Management for Dense Maps on Memory-Constrained Robots

Abstract: Running a language model and a dense 3D map on one device means the two share one memory, and exceeding its limit is an out-of-memory failure. Map systems that keep a bounded working set bound it by a count of keyframes, but the memory one keyframe occupies varies widely with what it carries, so a count does not control what the working set occupies. The unit of admission is as poorly matched, since a radius around the pose, a fixed cell, or the graph neighbors of the last matched keyframe all move parts of the map that are not needed. We therefore propose two decisions. First, the memory the resident map uses is bounded directly, and the cost of every admission is known before any data is read, from per-keyframe resident sizes computed offline and carried with the map. Second, keyframes are admitted and released at a semantic spatial unit, the rooms and corridors of the building, so that the robot holds the zone it occupies together with the zone it is about to enter while the zones it has passed are released. Every admission measures the resident total, makes room, and only then reads, so the budget is never overshot. On two NVIDIA Jetson boards the budget holds in every run and the model gives up at most 0.8 percent of its throughput, whereas without it the operating system kills the model or the map. Under that budget, and over the environments evaluated, admission by semantic zone localizes 6-9 points more of the query frames than a geometric partition of matched zone count and about twice as much as the retrieval of RTAB-Map, at a lower ATE. Bounding memory lets the map and the model coexist, and moving the map in semantic zones makes the bounded map usable.

Sat 24 OctRobotics
The gist
When small robots try to run both a language model and keep a detailed 3D map, they can run out of memory because both use the same storage space. The authors propose a way to carefully control how much memory the map uses by limiting it directly instead of just limiting the number of map snapshots, which can vary greatly in size. They group map information by meaningful spaces like rooms and corridors, keeping only the areas the robot is in or about to enter. This lets the robot avoid crashes and still localize itself more accurately compared to other methods.
Open → 2610.33574v1

Compact lidar model improves robot navigation in crowded environments

RangeNav: Task-Aligned LiDAR Range-Image Pretraining for Robot Navigation

Abstract: Autonomous mobile robots operating in dense crowds and cluttered sites require $360^\circ$ perception to react to surrounding hazards. While panoramic LiDAR offers full-surround coverage, learning representations within reinforcement learning (RL) is sample-inefficient under sparse rewards, and existing 3D pretrained point-cloud encoders are too memory-heavy for massively parallel RL simulation. We present RangeNav, a compact (183k parameters) LiDAR range-image encoder pretrained entirely on synthetic data. RangeNav uses anisotropic convolutions and circular padding to handle the extreme aspect ratio and azimuthal seam of range images. We pretrain the network using a navigation-aligned objective combining range reconstruction, semantic segmentation, and bird's-eye-view (BEV) prediction, then freeze it for policy learning. In simulation, adding semantic and BEV supervision increases success rates over a reconstruction-only baseline by an average of 32.03 and 23.73 percentage points (pp) in procedurally generated training and out-of-distribution realistic test environments, respectively. Zero-shot real-world deployments on a Unitree B2-W quadruped demonstrate robust navigation through dense crowds, low-light settings, and construction zones where limited-field-of-view depth policies struggle. Code and models will be open-sourced.

Wed 21 OctRoboticsMachine Learning
The gist
Robots moving through crowded or complex places need to see all around them to avoid obstacles. The researchers created RangeNav, a small and efficient computer program that helps robots understand 360-degree lidar images. They trained RangeNav using fake data to recognize distances, object types, and overhead views, which improves robot navigation success. This new method works better than previous approaches and even performs well in real-world tests with a robot dog going through crowds and tricky spots.
Open → 2610.29845v1

Real time object aware vSLAM improves navigation in dynamic scenes

CAOS-SLAM: Class-Agnostic Object-Aware Real-Time vSLAM for Dynamic Environments

Abstract: Robust visual Simultaneous Localization and Mapping (vSLAM) in dynamic environments faces challenges due to the presence of moving objects that may introduce erroneous observations, corrupt the map reconstruction and overall degrade localization accuracy. Existing vSLAM approaches often rely on predefined semantic classes or costly pixel-wise classification, limiting their robustness or applicability in real-world scenarios. This paper presents CAOS-SLAM, a dynamic vSLAM framework that extends ORB-SLAM3 with a hybrid perception pipeline to detect, track, and filter both semantically known and previously unseen moving objects in real-time. On-demand panoptic segmentation is performed for object discovery, with candidate unknown objects instantiated through depth-based clustering of unclassified regions. Between panoptic updates, persistent object tracks are updated using lightweight detection when available or temporal mask propagation based on world-frame Kalman prediction, depth consistency, and appearance similarity. The estimated object motion is then used to reject dynamic features from the SLAM pipeline. Experimental results on multiple dynamic RGB-D datasets and real-world experiments demonstrate improved localization accuracy over state-of-the-art dynamic vSLAM approaches, achieving up to 59% lower localization error in scenes characterized by unknown dynamic objects and up to 88% lower error in heavily occluded scenes, while achieving a median processing rate of 30 FPS on a laptop. Furthermore, the tracking module provides estimates of object trajectories and velocities, supporting potential future applications in perception, navigation, and motion-planning tasks in dynamic environments. Code and dataset will be released upon acceptance.

Mon 19 OctRobotics
The gist
Visual SLAM systems help robots and devices understand where they are by mapping their surroundings using cameras. Moving objects like people or cars can confuse these systems, causing mistakes in mapping and positioning. The authors created CAOS-SLAM, which can detect both known and unknown moving objects without needing prior labels, using a mix of visual perception techniques. This helps the system ignore moving objects when mapping, making localization much more accurate and faster, even on regular laptops.
Open → 2610.26633v1

Passive-wheel robots learn efficient skating to navigate cluttered spaces

SkateNav: Learning Energy-Efficient Navigation through Adaptive Skating on Passive-Wheel Quadrupedal Robots

Abstract: Passive-wheel quadrupedal robots offer energy-efficient locomotion through leg-driven propulsion and momentum-driven gliding. However, navigation requires coordinating steering with intermittent propulsion and adapting push-glide timing to task demands. Velocity-tracking objectives may suppress the speed variations inherent in skating, discouraging efficient gliding. To address these challenges, we propose SkateNav, a unified skating-navigation framework based on reinforcement learning. A time-budgeted goal-reaching objective allows the policy to jointly learn steering, propulsion, and their temporal allocation. An adaptive skating cycle structures gliding, pushing, and recovery, while frequency modulation at every control step adjusts phase durations to navigation demands. Simulation experiments demonstrate higher average success rates and lower cost of transport than the evaluated planner-controller baselines, with ablations confirming the contributions of cycle structure and frequency adaptation. Real-world experiments demonstrate zero-shot transfer from simulation to a passive-wheel Mini Cheetah across cluttered indoor spaces, slopes, and paved outdoor surfaces. Supplementary videos are available on the project page: https://ktw1404.github.io/SkateNav/

Mon 19 OctRobotics
The gist
Navigating environments with passive-wheel quadrupedal robots is tricky because these robots move by pushing and gliding, which requires careful timing and steering. The authors developed SkateNav, a method that helps the robot learn how to steer and decide when to push or glide to reach goals efficiently. Their approach uses reinforcement learning and adapts the robot’s movement cycle based on the task, improving success and saving energy compared to other methods. They tested it both in simulations and with a real robot across various indoor and outdoor areas without additional training.
Open → 2610.25962v1

Robots improve 3D sensing with folding wheels and sensor fusion

Enhancing Robotic Perception and Adaptability through Sensor Fusion and Origami-Inspired Designs

Abstract: Compact mobile robots must recover scene geometry under changing lighting and surface texture while working within tight payload and cost limits. We present a compact mobile robot that uses origami-inspired wheels for locomotion and active control of its sensing geometry. As the wheels move between terrain-adaptive configurations, the changing chassis pitch sweeps a 2D LiDAR through intermediate elevations; held wheel positions provide a chosen viewing angle. An IMU accounts for chassis attitude, and a fusion node projects LiDAR returns into the RGB-D depth stream supplied to RTAB-Map. The arrangement uses the wheel actuation already present on a sub-300 USD, sub-2 kg prototype to extend the scanner's viewing geometry. We assess depth fusion in a textureless indoor corridor and an outdoor sunlit area, with three runs per sensor configuration in each setting. Mean full-frame invalid-depth fractions fell from 21% to 11% indoors and from 48% to 18% outdoors. The prototype combines improved depth coverage with a continuously adjustable LiDAR viewpoint using the same actuation that reconfigures its wheels.

Wed 7 OctRobotics
The gist
Robots need good depth sensing to understand their surroundings, but changing light and smooth surfaces make this hard. The authors created a small robot with foldable wheels, inspired by origami, that lets the robot scan its environment from different heights and angles without extra parts. This clever use of existing moving wheels, combined with combining data from different sensors, helped the robot capture better 3D images indoors and outdoors. Their tests showed fewer missing depth points, making the robot’s vision more reliable in tricky environments.
Open → 2610.09828v1

Hybrid planner learns task and motion plans from few real demos

Learning Task and Motion Plans from Real Demonstrations with Hybrid Flow Matching

Abstract: Long-horizon mobile manipulation requires a task plan and the motion that executes it. Generative planners trained on demonstration produce both in one pass, requiring neither a symbolic domain nor search. To date, however, they have relied on thousands of scripted demonstrations of fixed-base arms and executed open loop. This paper presents a hybrid flow matching planner: a single network generates the symbolic plan with masked discrete flow matching and the motion trajectory with continuous flow matching. Unlike prior generative planners, we aim to learn from a much smaller set of demonstrations and to execute the plan in closed loop. Two properties of the data compensate for the small dataset. A demonstration resumed from any of its intermediate actions is itself a demonstration, which multiplies the training samples and enables replanning after every action. Objects of the same kind are interchangeable, which turns demonstrations of one goal into demonstrations of every permuted goal. Our base implementation produces valid plans on 68% of held-out scenes; a training and generation scheme for the discrete plan raises this to 76%, and replanning after every action raises the task completion rate from 40% to 53% in a kinematic simulation. The planner matches the task completion rate of motion-only flow matching policies while additionally providing the symbolic plan, and it outperforms previous hybrid diffusion formulations on both task completion and plan validity. We validate the planner on a real mobile manipulator. Videos and project page: https://andreumatoses.github.io/research/hybrid-flow-planning

Sat 3 OctRoboticsMachine Learning
The gist
Planning how a robot moves and what tasks it performs usually needs a lot of examples or complicated programming. The authors created a method where one neural network learns both the steps a robot should take and the motions to do them, using only a small set of real demonstrations. They improved training by treating parts of demonstrations as new examples and swapping similar objects during learning. Their system works better than previous ones, can plan and adjust as it goes, and was tested successfully on a real mobile robot.
Open → 2610.04771v1

Video transformers trimmed to speed up collision prediction with little accuracy loss

Investigating Spatiotemporal Redundancy in Video Transformer for Collision Anticipation

Abstract: In worker-equipment proximity monitoring, video transformers are widely used for collision anticipation and have demonstrated strong performance. However, their accuracy comes with substantial computational demands, creating a tension with the need for low-latency inference on mobile robots and the pursuit of lower-carbon computation in construction. To address this, this study investigates where computation within an established video transformer is redundant and whether that redundancy can be removed without materially degrading predictive performance. Using VideoMAEv2-Base on the Nexar Collision Prediction dataset, we first examine how collision-relevant information evolves across network depth and then investigate two complementary forms of redundancy: structured capacity redundancy in multilayer perceptrons (MLPs) and spatiotemporal redundancy in the token stream. Linear probes show that interpretable motion cues, including flow magnitude, looming, and approach versus retreat, are most accessible at intermediate layers, whereas collision-label discrimination strengthens toward the final layer. Token redundancy is axis-specific: adjacent temporal-token similarity reaches 0.970 in later layers, while spatial similarity falls to 0.297, indicating substantially greater redundancy across time than across space. Exploiting this asymmetry, temporal token merging reduces backbone computation from 356.99 to 178.50 GFLOPs and latency from 12.69 to 7.21 ms per clip, a 1.76x speedup, while mean average precision changes only from 0.7478 to 0.7443. Importance-guided retention of 50% of MLP units preserves an AUC of 0.753, compared with 0.529 under matched random retention, and reveals that pruning alters score calibration before discriminative ranking collapses. These findings establish a new pathway for pursuing faster algorithms through targeted temporal token compression and neuron pruning.

Sat 3 OctComputer Vision and Pattern Recognition
The gist
Collision anticipation systems use video transformers to watch workers and equipment and predict accidents, but these systems can be slow and costly to run. The authors studied where in the video transformer the computer is doing extra work that doesn’t improve predictions much. They found that many video frames over time are very similar and can be combined without losing much accuracy. Also, parts of the model’s neurons can be removed carefully to keep performance strong. These tricks together make the system almost twice as fast without hurting its ability to predict collisions.
Open → 2610.04727v1

Edge spiking neural networks learn locally and in parallel during sensing

Learning While Inferring: Local and Parallel Learning for Edge SNNs across Sensing Modalities

Abstract: Edge intelligence requires models to sense continuously in real time and to keep adapting on-device, all under tight compute, energy, and memory budgets. Although spiking neural networks (SNNs) enable efficient event-driven inference, standard surrogate-gradient backpropagation (BP) serializes updates and blocks ongoing inference. We investigate Bidirectional Spike-Based Distillation (BSD) as an on-device learning principle that lets edge SNNs learn while inferring. BSD couples a stimulus-driven forward network with an independent target-driven reverse network and aligns their intermediate representations through local objectives. Because the two pathways have disjoint computation graphs until alignment, forward inference, reverse inference, and stage-wise updates can run concurrently. On 25 benchmarks spanning the five sensing modalities of SOUL, BSD stays within 3.8 percentage points of matched BP baselines on average, while only its forward branch is needed at deployment. The learned representations also transfer well to few-shot class-incremental learning without replaying past data. By removing the dense floating-point backward chain, BSD reduces the projected training latency to $0.72\times$ and the estimated training energy to $0.36\times$ that of BP.

Fri 2 OctNeural and Evolutionary Computing
The gist
Devices that process sensory data close to where it's collected need to learn and adapt continually without using too much power or memory. The authors explored a new way for spiking neural networks to update their learning while still making predictions, unlike standard methods that have to pause to learn. Their approach uses two separate networks working at the same time to teach each other locally, letting the device learn continuously and efficiently. This method performs almost as well as traditional ones, uses less energy during training, and can adapt to new tasks with less data.
Open → 2610.03149v1

Robot system improves object search by learning spatial and semantic context

LOCUS: Landmark-Oriented Container Discrimination Using Spatial Graphs

Abstract: As robots are increasingly deployed in unstructured, real-world environments, the ability to reason about complex spatial and semantic relationships among objects remains a fundamental challenge in enabling robust and generalizable manipulation and navigation. For example, deciding where to search for an object that is not in plain sight depends on where it is physically plausible as well as semantically likely. Popular methods such as cosine similarity with CLIP struggle to disambiguate between spatially and semantically similar objects that could contain a target object. We propose Landmark-Oriented Container Discrimination Using Spatial Graphs (LOCUS). To jointly reason about spatial information and semantics, we train a GNN to update CLIP embeddings on a full environment scene graph by passing embedding information between nodes based on proximity. CLIP embeddings, augmented by semantic knowledge from a fusion of household ontologies provide a robust semantic signal. We evaluate in simulation on a noiseless scene graph from simulator metadata, making the approach detection-agnostic. Our approach outperforms random, CLIP, Tidybot, and an LLM planner in the majority of room classes and configurations in simulation. To show our approach is robust to scene graph node placement and label noise, we demonstrate a physical mobile manipulator running in an exploration pipeline start-to-finish including scene graph labels from Detic.

Fri 2 OctRobotics
The gist
Finding objects hidden around a home is hard because many containers look and seem alike. The authors created a new method called LOCUS that helps robots figure out where items might be by considering both where containers are placed and what they usually hold. They use a special graph-based neural network to mix location and meaning information, making the robot smarter at guessing the right container. They tested it in simulations and on a real robot, showing better results than some popular methods.
Open → 2610.02803v1

Fisheye camera detection improved with minimal retraining steps

FUSEye: Training-Light Fisheye Detection with Overlapping Views and Zero-Initialized Adapters

Abstract: Fisheye cameras give mobile robots a single-sensor, low-cost view of their surroundings, yet the COCO-pretrained detectors that practitioners routinely reuse fail on them: strong radial distortion warps local image structure, while boundary compression shrinks objects to near-invisible sizes. Full fine-tuning closes much of the gap but requires abundant fisheye labels and compute. We present FUSEye, a training-light framework that turns a frozen-backbone COCO-pretrained extra-large YOLO26 detector (YOLO26-x) into a fisheye detector. FUSEye adds roughly 227k new parameters while updating the inserted modules and the pretrained detection head. It addresses the transfer gap at three causally linked levels. At the input level, overlapping grid view generation and box remapping (GridViews) enlarge compressed boundary regions. At the feature level, zero-initialized residual adapters (Z-Adapters) correct distortion-induced feature misalignment. At the decision level, learned cross-projection agreement fusion (AgreeFusion) promotes low-confidence detections only when they are supported by consistent evidence across multiple views. On the WoodScape surround-view fisheye benchmark, FUSEye raises YOLO26-x from 0.148 to 0.266 mAP50 and retains 84.3% fully fine-tuned accuracy. Moreover, randomly using only 25% of the labeled training images, FUSEye achieves 0.2597 mAP50, retaining 97.6% of its full-label performance. FUSEye also consistently improves YOLOv8-11 detectors, showing that the recipe is architecture-agnostic. Source code will be available at https://github.com/Su-wenya/FUSEye.

Fri 2 OctComputer Vision and Pattern RecognitionRobotics
The gist
Detecting objects in images from fisheye cameras is hard because the curved lens distorts what the camera sees, especially near the edges. The authors propose FUSEye, a method that tweaks a popular object detector with only a small number of new parts and training steps to handle these distortions better. They improve detection accuracy significantly, even when using much less labeled training data. Their approach works across different detection models and keeps the original detector mostly unchanged.
Open → 2610.02799v1

Robust safety guarantees for nonholonomic robots navigating active obstacles

Adversarially Robust Geometric Safety Certificates for Nonholonomic Robots Against Maneuvering Obstacles

Abstract: Safe navigation against obstacles that can actively maneuver within bounded capabilities remains challenging: robust control barrier function methods typically treat obstacle actions as generic disturbances, while differential-game approaches are computationally expensive for online navigation. We propose an adversarially robust geometric certificate that accounts for the worst-case effect of admissible obstacle maneuvers directly in the safe-set geometry through a closed-form contraction of the certificate parameters. The construction exploits a structural property of line-of-sight (LoS) certificates: the robot and obstacle actions enter the certificate through a common state-dependent geometric gain. This gain cancels in the worst-case comparison, reducing the differential game to a direct comparison between obstacle maneuvering capability and the weaker of the robot's longitudinal and steering authorities. Instantiated on the parabolic certificate, the construction yields Adversarially Robust Dynamic Parabolic Control Barrier Functions (AR-DPCBF), for which we establish sufficient conditions for forward invariance of the contracted safe set against all admissible obstacle maneuvers under kinematic bicycle dynamics with bounded inputs. When the obstacle capability is unknown, a sliding-window estimator supplies a high-probability upper bound, allowing the guarantee to be retained with the corresponding coverage probability. We further formulate soft and buffered variants to recover feasibility in dense environments. Simulations across obstacle capabilities, densities, and capability mismatch show substantial reductions in barrier violations and collisions and demonstrate that pointwise robustification of the barrier derivative cannot substitute for contraction of its geometry.

Tue 29 SeptRoboticsMultiagent Systems
The gist
Keeping robots safe when moving around obstacles that can actively move is tough. The authors developed a new mathematical way to define safe zones that consider the worst actions obstacles might take. Their method simplifies complex calculations by focusing on a specific geometric property, making it easier and more reliable for robots with steering limits to avoid collisions. They also include a way to estimate unknown obstacle abilities and adjust safety zones accordingly. Tests with simulations show that their approach reduces accidents better than other safety methods.
Open → 2609.37126v1

UAV memories combined on ground to improve low-altitude answers

Memory in the Sky: Low-Altitude Question Answering with Multi-Agent Memory Aggregation

Abstract: This paper studies low-altitude question answering (LAQA), in which distributed unmanned aerial vehicle (UAV) memories are aggregated at a ground server to answer questions about observations over a long horizon. Unlike conventional resource allocation based on sensing, communication, control, or computation metrics, LAQA requires an explicit measure of memory value. We propose a generative adversarial exam (GAE) that uses forward simulation to evaluate memory retrieval and exam scores to quantify memory quality. This enables the downstream QA value of candidate memories to be measured and optimized without accessing the internal mechanisms of the black-box captioning, retrieval, and reasoning pipeline. Building on this metric, we develop a memory-centric (MemCen) framework that jointly selects UAVs and allocates transmit power to maximize memory quality under communication constraints. In the noise-limited regime, we derive a QoM-aware capped water-filling law that explicitly connects task utility with physical-layer power allocation. We further develop penalty successive optimization (PSO) and learning to memorize (L2M) solvers. MemCen achieves QA accuracies of 92.4% and 84.0% in CARLA Town04 and Town05 under static and dynamic communication conditions, respectively. In real-world experiments, MemCen achieves 88.5% QA accuracy on the panoramic multi-agent system (PMAS) benchmark. Finally, UAV-to-robot-dog demonstrations further validate the practical utility of the acquired memories for environmental understanding and navigation.

Mon 28 SeptRoboticsInformation Theory
The gist
Answering questions about what drones see over time is hard because their memories are spread out and communication can be limited. The authors propose a way to score and pick the most useful drone memories to send to a ground server, even without looking inside complex AI systems. They also develop methods to pick which drones to use and how much power to send data using new formulas. Their approach works well in both simulated and real-world tests, helping robots understand and navigate environments better.
Open → 2609.35431v1

Panoramic vision improves robot navigation with longer planning

PanoVLN: Towards Effective Panoramic Vision-and-Language Navigation

Abstract: Recent vision-language models (VLMs) have advanced vision-and-language navigation (VLN), enabling models to predict navigation actions from visual observations and language instructions. In this work, we explore VLN with panoramic observations and introduce PanoVLN. The motivation is straightforward: more complete visual context should enable better-informed navigation decisions. For example, a panorama can reveal a passage outside a perspective camera's field of view, allowing the model to identify the intended route without additional exploration. However, we find that simply replacing perspective images with panoramas yields only limited gains. Our diagnosis suggests that fully exploiting wider visibility requires modifications to action prediction, training supervision, and visual representation. First, wider visibility supports longer-horizon action planning. We make the model predict longer action sequences, enabling larger turns and subsequent movement from a single panorama. Specifically, we introduce a confidence-guided execution (CGE) strategy that dynamically determines how many predicted actions to execute before replanning. Second, wider visibility also brings more complex route choices. We therefore construct training routes with frequent branching points and clear instructions to provide targeted supervision for route selection. Third, panoramic navigation requires understanding spatial relationships across viewing directions, beyond recognizing individual landmarks. We combine semantic and geometric features from RGB panoramas to capture both scene content and spatial layout without adding visual tokens. With a 4B backbone and RGB-only input, PanoVLN surpasses the previous SOTA by 11.9% and 8.7% in success rate on R2R-CE and RxR-CE Val-Unseen. Real-world experiments on a quadruped further demonstrate faster navigation with fewer pauses than prior VLN methods.

Mon 28 SeptComputer Vision and Pattern RecognitionRobotics
The gist
The authors studied how robots can better follow instructions in indoor spaces by looking around more completely using panoramic views instead of narrow camera shots. They found that just having a wider view isn't enough; the robot also needs to plan longer sequences of moves, be trained with routes that have more choices, and understand the spatial layout better. They built a method called PanoVLN that combines these ideas and tested it in various tasks, showing it helps robots navigate more successfully and efficiently. Their experiments included real-world runs on a robot dog, which moved faster and paused less than previous methods.
Open → 2609.34759v1

SparseNav improves indoor robot navigation with selective map labeling

SparseNav: Instruction-conditioned Sparse Semantic Perception for Training-Free Vision-Language Navigation

Abstract: Map-based vision-language navigation (VLN) relies on persistent spatial representations to connect language understanding with geometric planning. However, acquiring semantics beyond the needs of the current instruction can introduce unnecessary perception cost and irrelevant annotations. Continuously accumulating unrelated objects may not only waste computation, but also clutter the visual-spatial representation consumed by the vision-language model (VLM) planner. To address this problem, we present SparseNav, a training-free framework that follows a less-is-more principle for semantic navigation. SparseNav persistently maintains a lightweight geometric bird's-eye-view (BEV) map and sparse landmark memory, acquiring new semantics on demand using the active sub-instruction to decide what is worth grounding. An instruction manager first tracks navigation progress and identifies the active landmark query. An instruction-conditioned perception mechanism then invokes open-vocabulary segmentation when the queried landmark is visible and its metric location can inform the next decision. The resulting landmark memory supports VLM selection among hybrid frontier and local directional waypoint candidates. Without any additional training, SparseNav achieves success rates of 42.8% on R2R-CE and 40.7% on RxR-CE, both on the Val-Unseen splits. Controlled ablations examine semantic perception strategies and the contributions of individual framework components. Furthermore, we successfully deployed SparseNav on a Unitree Go2 quadruped equipped with an Intel RealSense D455 RGB-D camera for geometric mapping and landmark grounding and a Livox MID-360 LiDAR for localization, without a prebuilt map. We validated its effectiveness across multiple indoor environments using instruction-conditioned waypoint navigation.

Tue 22 SeptRobotics
The gist
Indoor robots need to understand both language instructions and their surroundings to find their way. The authors created SparseNav, a system that only adds important landmarks to a map based on the current instruction, saving computing power and avoiding confusion. It works without extra training and was tested successfully in simulated environments and on a real robot navigating indoors. This selective approach helps robots follow complex directions more efficiently.
Open → 2609.26408v1

Omnidirectional walking robot improves safety and energy use in obstacle tasks

Safety-Constrained Model Predictive Control for an Omnidirectional Walking Assistive Robot Using Control Barrier Function

Abstract: Providing safe and effective mobility assistance plays a crucial role in restoring independence and enhancing the quality of life for individuals with motor impairments. In this context, robotic walking assistive devices have recently emerged as promising solutions to provide physically compliant interaction while ensuring user safety and support. This paper presents a novel control framework for an omnidirectional Walking Assistive Robot (I-WANDER) that integrates a Control Barrier Function (CBF) formulation into a Model Predictive Control (MPC) scheme to explicitly enforce collision-avoidance safety constraints while optimizing for energy efficiency and smooth human-robot collaboration. The method was experimentally evaluated with 12 healthy participants performing two different walking tasks using both the proposed CBF-based MPC controller (CB-MPC) and a variable admittance controller (AC). The first task involved structured navigation through a U-shaped corridor, whereas the second consisted of a single-obstacle avoidance task performed blindfolded to ensure the obstacle was unexpected. Comparative results show that the CB-MPC architecture significantly reduces energy consumption and mechanical work (p < 0.01) without compromising motion smoothness, while also decreasing the number of obstacle collisions. Overall, the findings highlight the potential of the proposed control architecture to enhance both safety and efficiency in robotic walking assistance.

Tue 22 SeptRobotics
The gist
Helping people with motor difficulties move safely and comfortably is important. The authors designed a robot that helps users walk in any direction, using smart controls to avoid bumping into things and to save energy. They tested their new control system with people walking through tricky paths and avoiding unexpected objects. Results showed this method lowered energy use and collisions while keeping smooth movement, suggesting a safer, more efficient walking aid.
Open → 2609.25994v1

Foundation model confidence and depth improve detection of unknown objects

Combining Foundation Model Confidence and Monocular Depth for Training-Free Out-of-Distribution Segmentation

Abstract: Autonomous vehicles operating in open-world scenarios are inevitably confronted with previously unknown objects, such as exotic animals or loose cargo. The reliable detection and segmentation of these out-of-distribution (OOD) objects is therefore crucial for a safe understanding of the environment and decision-making. Most existing approaches require access to OOD training samples, retraining of the segmentation backbone, or dedicated auxiliary architectures, limiting their practical applicability. We propose a training-free method that derives dense OOD scores directly from the confidence predictions of a foundation segmentation model, without any task-specific fine-tuning or access to anomalous data. To improve the robustness of our OOD segmentation, geometric information from monocular depth estimation is incorporated into the decision process, providing complementary cues to uncertainty-based predictions. We evaluate the proposed method on the SegmentMeIfYouCan benchmark and additionally assess its performance on OOD tracking in video sequences, reflecting the temporal nature of real-world perception systems. The method performs strongly on road-centered benchmarks.

Sat 19 SeptComputer Vision and Pattern Recognition
The gist
Self-driving cars need to notice when they see things they haven't seen before, like strange animals or loose items on the road, to stay safe. The authors propose a way to detect these unusual objects without needing extra training or special data. They combine how confident a big AI model is about what it sees with depth information from a single camera to spot these unknown things more reliably. Their method works well on road-focused tests and in video sequences, helping cars better understand their surroundings.
Open → 2609.22896v1

Robot navigation improves with smart waiting and rerouting decisions

SPARROW: Survival-POMCP for Adaptive Robot Routing, Observation, and Waiting

Abstract: Temporary obstacles that may block a robot's planned route create a sequential navigation problem: a robot must decide whether to wait for a blockage to clear, reroute, or acquire more information about the obstacle before acting. We formulate graph navigation among temporary obstacles as a partially observable semi-Markov decision process and introduce SPARROW, a belief-space planner built on Partially Observable Monte Carlo Planning (POMCP). SPARROW searches over traversal, observation, and finite-duration waiting actions while maintaining a particle belief over latent obstacle classes and clearance times. Class-conditioned survival models are learned online from both clearance observations and right-censored encounters where the robot reroutes before clearance is observed. A generative model simulates obstacle arrivals and clearances as each action unfolds, so the planner can account for blockages that may occur along alternative routes. We further introduce a value-of-learning criterion that trades the immediate cost of collecting labelled survival data against its expected reduction in future navigation regret. Across two simulation graphs and multiple obstacle-class settings, SPARROW reduces mean time-to-goal by 12-26% relative to OSCAR, a recent survival-based method for the same problem. On a physical mobile robot, SPARROW reduces mean time-to-goal by 20.5% relative to OSCAR while selectively observing, waiting, and rerouting as environment conditions change.

Thu 17 SeptRobotics
The gist
Robots often face temporary obstacles that block their way, like a door that might open soon or a hallway that might clear. The authors created a method called SPARROW that helps robots decide whether to wait, find a new path, or gather more information before moving. SPARROW uses a technique that plans while considering uncertain obstacles and learns how long they usually last. This approach helps robots reach their destinations faster both in simulations and real-life tests.
Open → 2609.21008v1

Robots navigate safer using semantics in 3D uncertainty maps

SemSafe-3DGS: Semantic Risk-Aware Active Navigation in Uncertain 3D Gaussian Splatting Maps

Abstract: Autonomous robots operating in partially observed environments must navigate safely while acquiring observations that improve future planning. Existing safety formulations generally reason primarily about geometry. Consequently, geometrically similar scene elements may induce comparable control responses despite having different semantic consequences. We present a semantic risk aware safe-active perception framework for navigation in attributed 3D Gaussian maps. Semantic attributes modulate an Average Value-at-Risk collision clearance model through class dependent risk weights, allowing safety-critical Gaussian primitives to receive greater influence in the composite barrier. The resulting weighted clearances are aggregated into a control barrier function, while a trajectory-relevant active perception barrier promotes observations that reduce geometric map uncertainty along the robot's anticipated motion. Both objectives are integrated in a unified CBF-QP that enforces semantic risk-aware collision avoidance as a hard constraint while relaxing information acquisition when it conflicts with safety or task progress. Experiments demonstrate efficient safety constraint, improved navigation through active perception, semantic dependent trajectory adaptation, and real-robot execution under Ackermann dynamics.

Wed 16 SeptRobotics
The gist
Robots need to move around safely even when they don’t see their whole environment clearly. This work helps robots understand what different parts of a scene mean, not just how they look or where they are. By giving more importance to objects that might cause more serious problems if hit, the robot can plan safer paths and also decide where to look next to improve its map. The authors tested their system on real robots that drive like cars and showed it helps avoid risky situations better while exploring.
Open → 2609.19330v1

Robot plans take humans intentions into account using vision language models

HINT-Plan: Human Intention-Aware Robot Task Planning in Context-Rich Environments using Vision Language Models

Abstract: Approaches to incorporating human awareness into mobile robot decision-making mainly focus on collision avoidance in low-level motion planning, often overlooking the challenges posed by human presence and high-level behavior. To address this vacancy, we present HINT-Plan, a novel approach to integrate human intention prediction into robot task planning. HINT-Plan employs Vision Language Models (VLMs) to anticipate high-level human intentions from third-person image observations, convert them into goal states, and solve joint task-planning problems. To effectively enable scene awareness in context-rich environments, we use hierarchical Scene Graphs (SGs) as high-level representations of the environment, and translate environmental topology and actionable knowledge into formal planning language to ensure executable plans. Evaluated in a photorealistic simulation, HINT-Plan achieves an overall success rate of 69.71% in joint human-robot task planning, substantially outperforming the baselines by up to 35.29%, while also reducing functional conflicts. The results show the effectiveness of explicitly incorporating inferred human intentions into formal multi-agent task planning for proactive human-aware robot decision-making.

Tue 15 SeptRoboticsArtificial Intelligence
The gist
Robots often avoid bumping into people but don't usually understand what people want to do around them. The authors developed HINT-Plan, which helps robots predict what a person nearby intends to do by looking at images and using special language-vision tools. The robot then plans its own tasks considering these human intentions and the environment details. This method was tested in a realistic simulation and showed better success in working alongside humans than previous approaches.
Open → 2609.17771v1

Neural control barrier functions enable safer robot navigation

VertexCBF: Improving Neural Control Barrier Functions via Vertex-Restricted Control Search

Abstract: As the number of autonomous robots continues to grow, safety becomes increasingly important. Control barrier functions (CBFs) provide a theoretically grounded framework for ensuring safety, but existing design methods often face limitations in effectiveness, scalability, or interpretability, and may result in overly conservative safe sets. In this paper, we propose \emph{VertexCBF}, a framework for learning neural CBFs in a scalable, systematic, and explainable way. We approximate the stationary Hamilton--Jacobi value function using a neural network trained via a combination of physics-informed and sparsely supervised learning. By exploiting control-affine dynamics and a convex polytope control set, under which the Hamiltonian is maximized at the control vertices, we efficiently generate supervision points via GPU-parallel vertex-restricted tree search, while a residual architecture guarantees that the learned CBF is never larger than the specified constraint function. We evaluate the method on 15 systems and compare it against relevant baselines, showing that it reliably recovers large safe sets where the baselines are conservative or fail completely. In addition, we perform a hardware experiment in which a mobile robot safely avoids pedestrians using a neural CBF trained with our method.

Fri 11 SeptRoboticsMachine Learning
The gist
Keeping robots safe is important as they become more common. Control barrier functions (CBFs) help robots avoid danger, but existing methods can be too cautious or hard to use. The authors created VertexCBF, a new approach that uses neural networks to better learn safe behaviors efficiently and understandably. Their method works on different robots, making safe zones larger and avoiding failures seen in older methods. They tested it on a real robot that safely avoided people walking nearby.
Open → 2609.12831v1

Pedestrian trajectory prediction improves using maps and awareness states

MamMA: A Mamba-Based Pedestrian Trajectory Prediction Algorithm Considering Occupancy Map and Pedestrian Awareness States

Abstract: Many pedestrian trajectory prediction algorithms have been proposed to improve the safety of navigation for mobile robots working in human-robot coexistence environments. Some pedestrian trajectory prediction algorithms extract information about obstacles near pedestrians from top-down view images to improve the accuracy of trajectory prediction. However, mobile robots typically create local occupancy maps using LiDAR, rather than top-down view images. Meanwhile, the vision sensors on board robots provide egocentric view images, which contain fine-grained behavioral information about the pedestrians near the robot. To better use the information collected by LiDAR and on-board vision sensors, we propose MamMA, a Mamba-based pedestrian trajectory prediction algorithm considering occupancy maps and pedestrian awareness states. MamMA divides the occupancy map by patches and extracts obstacle features from each patch to create map features. Pedestrian awareness states are divided and considered, as some studies show that awareness states affect the perception and speed of pedestrians. Furthermore, a Mamba-based model is proposed to predict the future trajectories of pedestrians based on different types of features. Experiments on the STCrowd, SiT, JRDB, ETH, and UCY datasets show that MamMA achieves better average displacement error and final displacement error than the state-of-the-art algorithms.

Mon 7 SeptComputer Vision and Pattern RecognitionMachine Learning
The gist
Predicting where people will walk helps robots move safely around them. The authors developed MamMA, a method that uses maps from laser sensors and also considers whether pedestrians are aware of their surroundings, which affects how they move. They break down the map into parts to better understand obstacles and use a Mamba-based model to predict pedestrian paths. Tests on several common datasets show that MamMA predicts movements more accurately than prior methods.
Open → 2609.08041v1

Robot improves person following using maps to find lost targets

Human-Aware Target Tracking and Navigation: Fusing Kinematic State Estimation with Structural Map Constraints

Abstract: Autonomous mobile robots performing person-following tasks often suffer from temporary occlusions and sensor track loss in dynamic environments. This research presents an end-to-end autonomous navigation stack that addresses target occlusion through map-informed spatial reasoning. The proposed system features a multi-modal perception pipeline, fusing deep learning-based visual tracking with 2-dimensional LiDAR point clustering to maintain high-fidelity tracking of a tagged person. A continuous state estimator integrates this perception data with wheel odometry and IMU sensors for stable localization. When the active track is lost due to occlusion, the system activates a map-based recovery framework. Leveraging a predefined topological map, the system executes a graph-based search to propagate the target's last known trajectory along structurally defined walking lanes, adhering to left-hand regional conventions. By generating a discrete set of feasible future trajectories, the robot reasons about potential structural trajectory changes, such as continuing a heading or turning at an intersection. This map-informed prediction is fed directly to the local obstacle avoidance planner, enabling the robot to continue following its target safely and predictably until the person is visually reacquired. Real-world evaluations in dense multi-person environments demonstrate the system's robustness, achieving a 71.4\% target reacquisition success rate during major occlusion events lasting up to 7 seconds.

Mon 7 SeptRobotics
The gist
Robots that follow people often lose track when the person is hidden behind obstacles like other people or walls. The authors created a system that helps the robot keep following by combining video and laser sensors with a map of the environment. When the robot loses sight, it uses the map to guess where the person might be and moves intelligently to find them again. This approach improved the robot’s success in re-finding people during real-world tests.
Open → 2609.07091v1