Unstructured navigational features, such as irregular planting or discontinuities, remain the primary failure mode for under-canopy agricultural robots. Existing geometric approaches often fail in these scenarios because they compress high-dimensional visual data into deterministic spatial references, effectively discarding the uncertainty and semantic context required to navigate ambiguous terrain. To address this, we present LeCropFollow, a visual navigation framework that bypasses explicit geometric modeling in favor of a learned latent representation. By integrating a self-supervised semantic heatmap extractor with TD-MPC2, a Model-Based Reinforcement Learning (MBRL) planner, our system optimizes trajectories directly within a latent manifold. The framework operates over the uncompressed heatmap signal, preserving the semantic context that geometric reductions discard. We demonstrate that this representational shift enables zero-shot transfer from simplified simulation to the physical world without fine-tuning. Extensive field experiments in late-stage corn fields show that LeCropFollow matches state-of-the-art baselines in unstructured rows but significantly outperforms them in plantation gaps, achieving a 2.4x reduction in semantic failures compared to keypoint-based methods. These results suggest that latent planning offers a robust alternative to geometric estimation for operations in heterogeneous agricultural environments. Code, models, and data available: https://felipe-tommaselli.github.io/lecropfollow .

Mobile robots are currently used in different applications and environments. Among the different types, two-wheeled balancing mobile robots (2WBMRs) represent a hybrid class of robotic systems that combine the versatility of biped locomotion with the efficiency and speed of wheeled motion. However, the complexity of the system dynamics and the need for efficient control strategies make the design and autonomy of such robots a significant challenge. This study presents the design of a dual-loop control with gain scheduling for the 2WBMR Ascento. The methodology comprises three main stages: deriving a simplified model assuming an inverted pendulum configuration, designing the balancing control for the torso, and regulating the variable height based on joint control. The results indicate that the proposed dual-loop control with gain scheduling reduces the pitch angle regulation error by 8.54% and the linear position tracking error by 16.77% compared to the fixed-gain strategy.

Offshore inspection and maintenance have increasingly been using legged robots for routine sensing, yet many useful interventions still require physical interaction with tools, containers, and task-relevant objects. Employing robots for these tasks can reduce operators' exposure in confined, elevated, or potentially explosive areas. This paper presents a language-guided grasping pipeline for a legged mobile manipulator operating under partial observation. An operator defines the target, the system grounds it in RGB with open-vocabulary detection and promptable segmentation, extracts an object-centric RGB-D point cloud, improves sparse geometry through depth compensation and point-cloud completion, and selects a 6-DoF grasp using collision, clearance, reachability, and approach constraints. The system is implemented on a quadruped robot with an arm and evaluated in two cluttered tabletop scenes motivated by small-object retrieval during inspection and maintenance. Across paired trials, the proposed pipeline achieved 9/10 successful grasps, compared with 3/10 for a view-dependent deployment baseline. In this controlled setting, object-centric completion and execution-aware selection reduced approach collisions and improved the reliability of language-guided grasping for supervised field manipulation.

Leader-follower interaction is an important paradigm in human-robot interaction (HRI). Yet, assigning roles in real time remains challenging for resource-constrained mobile and assistive robots. While large language models (LLMs) have shown promise for natural communication, their size and latency limit on-device deployment. Small language models (SLMs) offer a potential alternative, but their effectiveness for role classification in HRI has not been systematically evaluated. In this paper, we present a benchmark of SLMs for leader-follower communication, introducing a novel dataset derived from a published database and augmented with synthetic samples to capture interaction-specific dynamics. We investigate two adaptation strategies: prompt engineering and fine-tuning, studied under zero-shot and one-shot interaction modes, compared with an untrained baseline. Experiments with Qwen2.5-0.5B reveal that zero-shot fine-tuning achieves robust classification performance (86.66% accuracy) while maintaining low latency (22.2 ms per sample), significantly outperforming baseline and prompt-engineered approaches. However, results also indicate a performance degradation in one-shot modes, where increased context length challenges the model's architectural capacity. These findings demonstrate that fine-tuned SLMs provide an effective solution for direct role assignment, while highlighting critical trade-offs between dialogue complexity and classification reliability on the edge.

Accurate taxonomic identification of parasitoid wasps within the superfamily Ichneumonoidea is essential for biodiversity assessment, ecological monitoring, and biological control programs. However, morphological similarity, small body size, and fine-grained interspecific variation make manual identification labor-intensive and expertise-dependent. This study proposes a deep learning-based framework for the automated identification of Ichneumonoidea wasps using a YOLO-based architecture integrated with High-Resolution Class Activation Mapping (HiResCAM) to enhance interpretability. The proposed system simultaneously identifies wasp families from high-resolution images. The dataset comprises 3556 high-resolution images of Hymenoptera specimens. The taxonomic distribution is primarily concentrated among the families Ichneumonidae (n = 786), Braconidae (n = 648), Apidae (n = 466), and Vespidae (n = 460). Extensive experiments were conducted using a curated dataset, with model performance evaluated through precision, recall, F1 score, and accuracy. The results demonstrate high accuracy of over 96

Estimating traversability in unstructured environments requires conditioning on robot embodiment, as the same terrain can be traversable for one platform and unsafe for another. Existing methods often transfer predictions across morphologies through late-stage trajectory filtering rather than encoding platform constraints in the learned representation. We propose Capability-Aware Traversability (CAT), a framework that embeds physical limits directly into the spatial feature space. CAT grounds dense supervision masks in physical trajectories through an interactive annotation pipeline and modulates semantic terrain maps with robot-specific traversability vectors through Spatially-Adaptive Denormalization (SPADE) blocks. Across human-annotated and trajectory-aligned datasets, CAT leads all ranking-based metrics, improving AUROC by 11.0

Accurate taxonomic identification is the cornerstone of biodiversity monitoring and agricultural management, particularly for the hyper-diverse superfamily Ichneumonoidea. Comprising the families Ichneumonidae and Braconidae, these parasitoid wasps are ecologically critical for regulating insect populations, yet they remain one of the most taxonomically challenging groups due to their cryptic morphology and vast number of undescribed species. To address the scarcity of robust digital resources for these key groups, we present a curated image dataset designed to advance automated identification systems. The dataset contains 3,556 high-resolution images, primarily focused on Neotropical Ichneumonidae and Braconidae, while also including supplementary families such as Andrenidae, Apidae, Bethylidae, Chrysididae, Colletidae, Halictidae, Megachilidae, Pompilidae, and Vespidae to improve model robustness. Crucially, a subset of 1,739 images is annotated in COCO format, featuring multi-class bounding boxes for the full insect body, wing venation, and scale bars. This resource provides a foundation for developing computer vision models capable of identifying these families.

Perceptive legged locomotion over discontinuous terrain (e.g., stairs, gaps, and obstacles) requires adaptive behavior, as a single conservative gait cannot produce the anticipatory maneuvers needed for abrupt topology changes. Cast as multi-task reinforcement learning, this problem introduces a tension between sharing and separation. Tasks use a common locomotion base but have conflicting rewards, so a policy must share behavior while avoiding value interference. Prior work addresses only one side, with monolithic policies sacrificing specialization and hierarchical sub-policies sacrificing generalization across transitions and unseen terrain. We propose CTS-MoE, which combines a dense mixture-of-experts actor with perception-based gating to compose shared behaviors and a multi-critic with task-specific value heads to prevent interference. The model is trained end-to-end in a single-stage concurrent teacher-student setup that handles partial observability and avoids sequential distillation, with task labels used only during training. At deployment, routing depends solely on perception, allowing terrain adaptation without a high-level selector or terrain classifier. Experiments on a Unitree Go1 in simulation and on hardware across seen and unseen terrains show task-aware specialization, with lower tracking error and higher success rates than monolithic baselines. Project Website: https://cts-moe.github.io/ .

Collapsing terrains, often present in search and rescue missions or planetary exploration, pose significant challenges for quadruped robots. This paper introduces a robust locomotion framework for safe navigation over unstable surfaces by integrating terrain probing, load-bearing analysis, motion planning, and control strategies. Unlike traditional methods that rely on specialized sensors or external terrain mapping alone, our approach leverages joint measurements to assess terrain stability without hardware modifications. A Model Predictive Control (MPC) system optimizes robot motion, balancing stability and probing constraints, while a state machine coordinates terrain probing actions, enabling the robot to detect collapsible regions and dynamically adjust its footholds. Experimental results on custom-made collapsing platforms and rocky terrains demonstrate the framework's ability to traverse collapsing terrain while maintaining stability and prioritizing safety.

Teleoperating a robotic manipulator in industrial environments demands precision that camera-based interfaces alone struggle to deliver. The operator must align the end-effector with a target in clutter, under limited depth perception, and without colliding with the surrounding structures. This paper presents a shared-autonomy framework that assists the operator throughout this process. A single RGB-D camera captures the operator's arm motion and hand gestures without wearables, fiducials, or a calibration stage. The intended target is specified by a free-form text prompt, grounded by a vision-language model in the robot's gripper camera, and tracked across its onboard cameras by a promptable video-segmentation model, resulting in a grasp frame continuously separated from the obstacle map. Every commanded motion is executed by a GPU-accelerated model-predictive controller that enforces self- and environment-collision avoidance against an online volumetric reconstruction, while a potential field corrects the operator's reference toward the grounded target during the final approach. An autonomous mode can be gesture-triggered to complete the grasp on the same target without a separate perception pipeline. The framework is validated on a quadruped mobile manipulator. The interface achieves a positional RMSE of 59 mm relative to motion-capture ground truth, and the controller keeps the arm at least 18 cm from obstacles while the operator deliberately commands the arm into them by 6 cm. In an industrial valve manipulation and a pick-and-place task, the full framework succeeded in all trials, while ablating either the collision or the assistance module produced failures through complementary mechanisms, and autonomous execution succeeded in four of five trials per task.

Indoor robotics research increasingly relies on micro-UAVs whose airframe, electronics, and control software are fully open to modification. Off-the-shelf platforms rarely expose the low-level access required for such modifications, while building a custom alternative typically requires substantial engineering effort before flight testing can begin, leaving many laboratories to work within constraints that limit the scope of their research. We present MIRA (Modular Indoor Research Architecture), a low-cost, open-source micro-UAV for indoor research built around a replicable 3D-printed PLA airframe and a containerized low-level software package managing the companion-to-autopilot communication bridge via Micro XRCE-DDS. Designed as a white-box architecture, core subsystems are individually replaceable without firmware refactoring, supporting local fabrication and component substitution from existing lab inventory. We characterize MIRA through manual flight in position-control mode within an optical motion-capture volume, where the communication pipeline sustains a median companion-to-autopilot latency of 0.02 ms and power spectral density analysis confirms the structural vibration energy stays concentrated in a narrow 90 to 110 Hz band, isolated from the sub-20 Hz control bandwidth and within the autopilot safety thresholds.

Traditional on-policy reinforcement learning (RL) controllers for quadrupedal locomotion often suffer from low data efficiency, requiring millions of interactions with simulated environments to achieve stable control. We integrate model-based techniques that improve sample efficiency by augmenting PPO rollouts with synthetic data in a Dyna-style framework. Our method employs a learned transition model to generate short-horizon synthetic tails for each trajectory, anchored by physics-based simulation to preserve stability. A predefined scheduling strategy gradually integrates synthetic transitions, preventing model usage during early training stages when prediction accuracy is low. Through extensive ablation studies, we analyze how varying data parameters influence PPO's learning behavior. Finally, we validate our method in simulation on a Unitree Go1 robot, reaching convergence with substantially fewer simulation steps (19.64M vs. 27.53M) and a 12.24

Grey-box methods for system identification combine deep learning with physics-informed constraints, capturing complex dependencies while improving out-of-distribution generalization. Despite the growing importance of floating-base systems such as humanoids and quadrupeds, current grey-box models ignore their specific physical constraints. For instance, the inertia matrix is not only positive definite but also exhibits branch-induced sparsity and input independence. Moreover, the 6×6 composite spatial inertia of the floating base inherits properties of single-rigid-body inertia matrices. As we show, this includes the triangle inequality on the eigenvalues of the composite rotational inertia. To address the lack of physical consistency in deep learning models of floating-base systems, we introduce a parameterization of inertia matrices that satisfies all these constraints. Inspired by Deep Lagrangian Networks (DeLaN), we train neural networks to predict physically plausible inertia matrices that minimize inverse dynamics error under Lagrangian mechanics. For evaluation, we collected and released a dataset on multiple quadrupeds and humanoids. In these experiments, our Floating-Base Deep Lagrangian Networks (FeLaN) achieve better overall performance on both simulated and real robots, while providing greater physical interpretability.

Reinforcement learning has become the leading paradigm in legged locomotion, enabling complex behaviors from backflips to parkour through massively parallel simulation. Under PPO's non-stationarity, shallow networks remain the de facto architecture, supported by carefully staged curricula and environments, yet the representations these policies learn stay poorly understood, leaving no training-time signal of how they will behave on hardware. In this work, we empirically study locomotion policies through the effective rank of the policy Jacobian and show that conditioning rank on the gait phase exposes architectural structure that global rank averages away. In particular, we find that standard architectural choices, namely layer normalization and residual connections, allocate roughly two more dimensions of effective rank to swing than to stance, which is fully absent in vanilla MLPs. Building on this, we propose a simple recipe that turns these representational signatures into smoother, more reliable sim-to-real transfer. In practice, this results in roughly 3x lower joint jitter that holds from simulation onto a physical Spot, suggesting that representation health is an effective training-time lens to track sim-to-real smoothness.

This work presents a locomotion policy that enables quadruped robots to handle highly inclined open staircases. Such staircases are common in industrial environments and are characterized by large inter-step gaps. The proposed policy is trained using perception-based inputs obtained from ray-casting sensors and a dual-stage curriculum learning framework with carefully designed reward functions. Experimental results reveal an interesting emergent locomotion behavior that enables reliable stair ascent and descent while maintaining robustness across a range of climbing and descending velocities.

Soybean and cotton are major drivers of many countries’ agricultural sectors, offering substantial economic returns but also facing persistent challenges from volunteer plants and weeds that hamper sustainable management. Effectively controlling volunteer plants and weeds demands advanced recognition strategies that can identify these amidst complex crop canopies. While deep learning methods have demonstrated promising results for leaf-level detection and segmentation, existing datasets often fail to capture the complexity of real-world agricultural fields. To address this, we collected 640 high-resolution images from a commercial farm spanning multiple growth stages, weed pressures, and lighting variations. Each image is annotated at the leaf-instance level, with 7,221 soybean and 5,190 cotton leaves labeled via bounding boxes and segmentation masks, capturing overlapping foliage, small leaf size, and morphological similarities. We validate this dataset using YOLO11, demonstrating state-of-the-art performance in accurately identifying and segmenting overlapping foliage. Our publicly available dataset supports advanced applications such as selective herbicide spraying and pest monitoring and can foster more robust, data-driven strategies for soybean-cotton management.

The field of legged robotics has seen significant advances in recent years, with contact detection and state estimation emerging as essential components in achieving robust and adaptive locomotion. This review examines recent developments in contact detection algorithms for legged robots, including model-based, learning-based, and probabilistic approaches, each with unique strengths and limitations. Selecting an appropriate contact estimation strategy requires careful consideration of environmental uncertainties, sensors, and trade-offs between model interpretability and adaptability. Model-based methods excel in controlled environments by offering interpretable and efficient solutions, while probabilistic methods address uncertainties in noisy and dynamic settings, enhancing robustness. Meanwhile, learning-based methods demonstrate better adaptability and terrain generalization, making them ideal for complex and diverse environments. This review presents a comparative analysis of contact detection methods, highlighting their applicability and limitations. Finally, key challenges such as sensor fusion, computational complexity, and terrain adaptability are discussed to guide future research.

Quadrupedal robots are increasingly used for inspection and maintenance in dynamic environments such as offshore platforms, where the inherent oscillation of the environment emerges as a challenging condition to enable steady locomotion. In this context, this paper compares a model-based and a learning-based strategy for quadruped robot locomotion control operating in oscillating environments. The first approach consists of a Model Predictive Control (MPC) system, implemented in C++ with a ROS interface and simulated in Gazebo, combined with a low-level PD in joint space. The second approach uses a Deep Reinforcement Learning (DRL) framework, implemented and trained in Python using the Isaac Lab framework for robot learning, which optimizes our agent policy for steady locomotion. Our comparison focuses on evaluating the maximum oscillation frequency each strategy can handle before losing stability across three amplitudes: 0.3 m, 0.6 m, and 1.0 m. The model-based strategy demonstrates flawless stability up to 0.29 Hz, 0.16 Hz, and 0.10 Hz, respectively, but collapses abruptly beyond these thresholds. In contrast, the learning-based approach exhibits a more gradual performance decline, maintaining frequency tolerances of 0.77 Hz, 0.52 Hz, and 0.39 Hz. Furthermore, the learning-based method significantly reduces torque spikes, generally near 15 Nm compared to the +/- 30Nm fluctuations observed with the model-based strategy. Simulation results demonstrate the strengths and limitations of each strategy, highlighting key considerations for deploying four-legged robots in industrial inspection tasks on offshore platforms.