IROS 2024 Papers

Common deep learning models for 3D real-time environment perception often use pillarization/voxelization methods to convert point cloud data into pillars/voxels and then process it with a 2D/3D convolutional neural network (CNN). The pioneer work PointNet has been widely applied as a local feature descriptor, a fundamental component in deep learning models for 3D perception, to extract features of a point cloud. This is achieved by using a symmetric max-pooling operator which provides unique pillar/voxel features. However, by ignoring most of the points, the max-pooling operator causes an information loss, which reduces the model performance. To address this issue, we propose a novel local feature descriptor, mini-PointNetPlus, as an alternative for plug-and-play to PointNet. Our basic idea is to separately project the data points to the individual features considered, each leading to a permutation invariant. Thus, the proposed descriptor transforms an unordered point cloud to a stable order. The vanilla PointNet is proved to be a special case of our mini-PointNetPlus. Due to fully utilizing the features by the proposed descriptor, we demonstrate in experiment a considerable performance improvement for 3D perception.

Coordinating multiple robots is crucial for various real-life applications. Many Multi-Agent Path Finding (MAPF) algorithms have been proven to be successful in addressing this challenge. Nevertheless, some problems, such as unexpected delays during navigation, commonly arise when handling highlevel abstractions, potentially leading to failures or collisions in live executions. This paper proposes k-Robust Continuoustime Conflict-Based Search (kR-CCBS), a novel algorithm that overcomes some of these limitations. Our approach offers path planning with continuous time, leading to more precise routes than discrete time approaches. Additionally, we increase safety by incorporating k-robustness, enabling the system to adapt to agent failures due to delays and minimize collision risks. Comparative evaluations demonstrate that kR-CCBS outperforms similar works in effectiveness while maintaining reasonable costs, making it a promising solution for real-world multi-agent coordination scenarios.

This paper considers a Min-Max Multiple Traveling Salesman Problem (MTSP), where the goal is to find a set of tours, one for each agent, to collectively visit all the cities while minimizing the length of the longest tour. Though MTSP has been widely studied, obtaining near-optimal solutions for large-scale problems is still challenging due to its NP-hardness. Recent efforts in data-driven methods face challenges of the need for hard-to-obtain supervision and issues with high variance in gradient estimations, leading to slow convergence and highly sub-optimal solutions. We address these issues by reformulating MTSP as a bilevel optimization problem, using the concept of imperative learning (IL). This involves introducing an allocation network that decomposes the MTSP into multiple single-agent traveling salesman problems (TSPs). The longest tour from these TSP solutions is then used to self-supervise the allocation network, resulting in a new self-supervised, bilevel, end-to-end learning framework, which we refer to as imperative MTSP (iMTSP). Additionally, to tackle the high-variance gradient issues during the optimization, we introduce a control variate-based gradient estimation algorithm. Our experiments showed that these innovative designs enable our gradient estimator to converge 20× faster than the advanced reinforcement learning baseline, and find up to 80% shorter tour length compared with Google OR-Tools MTSP solver, especially in large-scale problems (e.g. 1000 cities and 15 agents).

Rapidly-exploring Random Trees (RRT) and its variations have emerged as a robust and efficient tool for finding collision-free paths in robotic systems. However, adding dynamic constraints makes the motion planning problem significantly harder, as it requires solving two-value boundary problems (computationally expensive) or propagating random control inputs (uninformative). Alternatively, Iterative Discontinuity Bounded A* (iDb-A*), introduced in our previous study, combines search and optimization iteratively. The search step connects short trajectories (motion primitives) while allowing a bounded discontinuity between the motion primitives, which is later repaired in the trajectory optimization step.Building upon these foundations, in this paper, we present iDb-RRT, a sampling-based kinodynamic motion planning algorithm that combines motion primitives and trajectory optimization within the RRT framework. iDb-RRT is probabilistically complete and can be implemented in forward or bidirectional mode. We have tested our algorithm across a benchmark suite comprising 30 problems, spanning 8 different systems, and shown that iDb-RRT can find solutions up to 10x faster than previous methods, especially in complex scenarios that require long trajectories or involve navigating through narrow passages.

The advent of tactile sensors in robotics has sparked many ideas on how robots can leverage direct contact measurements of their environment interactions to improve manipulation tasks. An important line of research in this regard is grasp force control, which aims to manipulate objects safely by limiting the amount of force exerted on the object. While prior works have either hand-modeled their force controllers, employed model-based approaches, or not shown sim-to-real transfer, we propose a model-free deep reinforcement learning approach trained in simulation and then transferred to the robot without further fine-tuning. We, therefore, present a simulation environment that produces realistic normal forces, which we use to train continuous force control policies. A detailed evaluation shows that the learned policy performs similarly or better than a hand-crafted baseline. Ablation studies prove that the proposed inductive bias and domain randomization facilitate sim-to-real transfer. Code, models, and supplementary videos are available on https://sites.google.com/view/rl-force-ctrl

Despite the impressive performance of Reinforcement Learning (RL), the black-box neural network backbone hinders users from trusting and deploying trained agents in real-world applications where safety is crucial. In order to make agents more trustworthy and controllable, for a given RL-trained policy, we propose to enhance its interpretability and make it human-controllable without retraining. We accomplish this goal by following a 3-step pipeline: 1) We interpret neurons by analyzing the causal effect of neurons on the kinematic attributes; To help agents unlock novel skills and enable human to assist agents in accomplishing tasks, 2) we locate the X-neuron, the optimal neuron that is capable of evoking the desired behavior; 3) and edit its activation values to achieve the precise control. We evaluate our method on various RL tasks ranging from autonomous driving to robot locomotion, and the results display that our approach outperforms previous work regarding almost all evaluation metrics. Through enhancing interpretability and introducing human control, the agents can improve safety and performance, even in unseen environments and novel tasks. For locomotion robots simply trained to walk forward, our method unlocks diverse controllable behaviors ranging from jump to backflip.

This paper explores the application of X-ray-guided magnetic fields for the wireless control of untethered magnetic robots (UMRs) within cerebral vascular phantoms. With a focus on addressing challenges associated with strokes and brain aneurysms, the study aims to enhance neurosurgical procedures by improving precision and maneuverability. Experimental findings showcase the feasibility and effectiveness of this innovative approach in navigating UMRs, characterized by a screw-shaped body and a ferromagnetic core, through complex vascular structures. Cone-beam computed tomography is employed to determine the tomography and provide various reference trajectories for the UMR inside the cerebral vascular phantom. Our motion control experiments show that the X-ray-guided magnetic fields enable the UMR to move along any intended path with an average success rate of 89%, allowing the UMR to move between the left and right common carotid artery to the left and right internal and external carotid artery.

We present placing via picking (PvP), a method to autonomously collect real-world demonstrations for a family of placing tasks in which objects must be manipulated to specific, contact-constrained locations. With PvP, we approach the collection of robotic object placement demonstrations by reversing the grasping process and exploiting the inherent symmetry of the pick and place problems. Specifically, we obtain placing demonstrations from a set of grasp sequences of objects initially located at their target placement locations. Our system can collect hundreds of demonstrations in contact-constrained environments without human intervention using two modules: compliant control for grasping and tactile regrasping. We train a policy directly from visual observations through behavioural cloning, using the autonomously-collected demonstrations. By doing so, the policy can generalize to object placement scenarios outside of the training environment without privileged information (e.g., placing a plate picked up from a table). We validate our approach in home robot scenarios that include dishwasher loading and table setting. Our approach yields robotic placing policies that outperform policies trained with kinesthetic teaching, both in terms of success rate and data efficiency, while requiring no human supervision.

This paper addresses the challenge of actuating millimetre-sized motors, which are wirelessly driven by external magnetic fields. Traditional approaches, relying on rotating magnetic fields, often inadvertently cause the entire robot – especially if it is small and lightweight – to rotate, instead of a specified shaft in the motor. To overcome this issue, our study introduces a novel mechanism that leverages symmetrically configured magnetic motors to cancel out the torques, thus preventing unwanted rotation of the robot. This is achieved by utilizing a magnetic field along a single axis to induce rotational movement. The design features two millimetre-sized rotating magnets that interact to achieve a 90◦ rotation, complemented by an external magnetic field that accomplishes the remaining 270◦, thus completing a full rotation. Furthermore, we demonstrate that applying a perpendicularly oriented magnetic field can inversely affect the motor’s rotation direction. A proof-of-concept experiment employing this mechanism successfully actuated a gripper in a water tank while it is free-floating, showcasing its potential for enhancing robotic applications at the sub-centimeter scale, where the small net torque of a miniature motor is essential.

We present WidthFormer, a novel transformer-based module to compute Bird’s-Eye-View (BEV) representations from multi-view cameras for real-time autonomous-driving applications. WidthFormer is computationally efficient, robust and does not require any special engineering effort to deploy. We first introduce a novel 3D positional encoding mechanism capable of accurately encapsulating 3D geometric information, which enables our model to compute high-quality BEV representations with only a single transformer decoder layer. This mechanism is also beneficial for existing sparse 3D object detectors. Inspired by the recently proposed works, we further improve our model’s efficiency by vertically compressing the image features when serving as attention keys and values, and then we develop two modules to compensate for potential information loss due to feature compression. Experimental evaluation on the widely-used nuScenes 3D object detection benchmark demonstrates that our method outperforms previous approaches across different 3D detection architectures. More importantly, our model is highly efficient. For example, when using 256 × 704 input images, it achieves 1.5 ms and 2.8 ms latency on NVIDIA 3090 GPU and Horizon Journey-5 computation solutions. Furthermore, WidthFormer also exhibits strong robustness to different degrees of camera perturbations. Our study offers valuable insights into the deployment of BEV transformation methods in real-world, complex road environments. Code is available at https://github.com/ChenhongyiYang/WidthFormer.

Recently, humanoid robots have made significant advances in their ability to perform challenging tasks due to the deployment of Reinforcement Learning (RL), however, the inherent complexity of humanoid robots, including the difficulty of designing complicated reward functions and training entire sophisticated systems, still poses a notable challenge. To conquer these challenges, after many iterations and in-depth investigations, we have meticulously developed a full-size humanoid robot, "Adam", whose innovative structural design greatly improves the efficiency and effectiveness of the imitation learning process. In addition, we have developed a novel imitation learning framework based on an adversarial motion prior, which applies not only to Adam but also to humanoid robots in general. Using the framework, Adam can exhibit unprecedented human-like characteristics in locomotion tasks. Our experimental results demonstrate that the proposed framework enables Adam to achieve human-comparable performance in complex locomotion tasks, marking the first time that human locomotion data has been used for imitation learning in a full-size humanoid robot. For more video demonstrations, please visit our YouTube channel: https://www.youtube.com/watch?v=7hK2ySYBa1I

In normal operations, when quadruped manipulators with impedance control experience external disturbances, they may become unstable and lose balance due to actuation saturation, affecting their stability, safety, and compliance with the environment. To address this issue, we propose a whole-body compliance controller to prevent unstable behaviors like slip, oscillation, and overshoot, which arise from actuation saturation. The controller includes an admittance scheme with a set-valued operator as the internal feedback, to constrain joint torques within actuators’ limits and ground reaction forces within friction cones to ensure stability against disturbances. Then, it formulates a hierarchical optimization problem using the Hierarchical Quadratic Programming (HQP) to impose the output of the admittance scheme while ensuring physical consistency to maintain compliance behaviors. Unlike traditional compliance control with one-dimensional torque limitations, our approach considers both joints torque limits of manipulator joints and friction cones of quadruped ground reaction as actuation saturation. This ensures overall compliance and stability for the quadruped manipulators, even under significant external forces, regardless of where they are exerted on the robot. We demonstrate through experiments involving variable stiffness environments and external forces during normal operations how effective our approach is in enhancing the safety of quadruped manipulators.

Wheel-legged robots combine the advantages of both wheeled robots and legged robots, offering versatile locomotion capabilities with excellent stability on challenging terrains and high efficiency on flat surfaces. However, existing wheel-legged robots typically have limited hip joint mobility compared to humans, while hip joint plays a crucial role in locomotion. In this paper, we introduce Whleaper, a novel 10-degree-of-freedom (DOF) bipedal wheeled robot, with 3 DOFs at the hip of each leg. Its humanoid joint design enables adaptable motion in complex scenarios, ensuring stability and flexibility. This paper introduces the details of Whleaper, with a focus on innovative mechanical design, control algorithms and system implementation. Firstly, stability stems from the increased DOFs at the hip, which expand the range of possible postures and improve the robot’s foot-ground contact. Secondly, the extra DOFs also augment its mobility. During walking or sliding, more complex movements can be adopted to execute obstacle avoidance tasks. Thirdly, we utilize two control algorithms to implement multimodal motion for walking and sliding. By controlling specific DOFs of the robot, we conducted a series of simulations and practical experiments, demonstrating that a high-DOF hip joint design can effectively enhance the stability and flexibility of wheel-legged robots. Whleaper shows its capability to perform actions such as squatting, obstacle avoidance sliding, and rapid turning in real-world scenarios.

Effective teleoperated avatar communication requires expressing social behaviors. Gaze behavior is one of the crucial social behaviors and includes reflexive reactions to the avatar’s surroundings and intentional responses to the operator’s speech and actions. Teleoperated avatars must have their gaze behavior controlled according to situational changes in both the avatar’s and operator’s contexts. However, it is not clear how to adjust the avatar’s gaze in response to changes in both situations. In this paper, we collect a dataset of gazing positions that the avatar is instructed to face, taking into account both avatar and operator situations, and annotation labels that represent both situations in detail. We then exploratorily analyze the ratio of gazing positions per situation through dynamic area-of-interest (AOI) analysis. Our analysis provides insights into determining the gaze behavior of teleoperated avatars.

Decentralized navigation methods rely primarily on local observations, lacking the global awareness needed to coordinate effectively within a multi-agent system. Exchanging relevant messages between agents can promote cooperation and improve navigation efficiency. We present a Reinforcement Learning (RL)-based decentralized navigation approach that learns ‘when,’ ‘what,’ and ‘with whom’ to communicate for safe and cooperative navigation. Our method leverages a visual transformer and self-attention mechanism to encode the local occupancy map and the state information of neighbors into fixed-length encodings, allowing it to handle an arbitrary number of neighbors for collision-free navigation. In addition, the network encodes the agent’s state information and observations of neighboring agents into a concise message vector by learning what information is crucial to communicate, which is shared with neighboring agents upon request. Moreover, to avoid indiscriminate broadcasting, the network learns when and with whom to communicate and request message vectors. Subsequently, the messages communicated alongside the local information are used to guide navigation decisions. We evaluate our method against state-of-the-art baselines in complex scenarios, including narrow corridors and environments with multiple agents. We observe considerable improvements in terms of navigation performance, showing up to ∼ 2× improvement in navigation success rates and a reduction of up to ∼ 20% in path length.

In this work, we present a control framework to effectively maneuver wheelchairs with a dynamically stable mobile manipulator. Wheelchairs are a type of nonholonomic cart system, maneuvering such systems with mobile manipulators (MM) is challenging mostly due to the following reasons: 1) These systems feature nonholonomic constraints and considerably varying inertial parameters that require online identification and adaptation. 2) These systems are widely used in human-centered environments, which demand the MM to operate in potentially crowded spaces while ensuring compliance for safe physical human-robot interaction (pHRI). We propose a control framework that plans whole-body motion based on quasi-static analysis to maneuver heavy nonholonomic carts while maintaining overall compliance. We validated our approach experimentally by maneuvering a wheelchair with a bimanual mobile manipulator, the CMU ballbot. The experiments demonstrate the proposed framework is able to track desired wheelchair velocity with loads varying from 11.8 kg to 79.4 kg at a maximum linear velocity of 0.45 m/s and angular velocity of 0.3 rad/s. Furthermore, we verified that the proposed method can generate human-like motion smoothness of the wheelchair while ensuring safe interactions with the environment.

Current methods for large-scale point cloud scene semantic segmentation rely on manually annotated dense point-wise labels, which are costly, labor-intensive, and prone to errors. Consequently, gathering point cloud scenes with billions of labeled points is impractical in real-world scenarios. In this paper, we introduce a novel weak supervision approach to semantically segment large-scale indoor scenes, requiring only 1‰ of the points to be labeled. Specifically, we develop an efficient point neighbor Transformer to capture the geometry of local point cloud patches. To address the quadratic complexity of self-attention computation in Transformers, particularly for large-scale point clouds, we propose approximating the self-attention matrix using low-rank and sparse decomposition. Building on the point neighbor Transformer as foundational blocks, we design a Low-rank Sparse Transformer Network (LST-Net) for weakly supervised large-scale point cloud scene semantic segmentation. Experimental results on two commonly used indoor point cloud scene segmentation benchmarks demonstrate that our model achieves performance comparable to those of both weakly supervised and fully supervised methods. Our code can be found in https://github.com/hhuang-code/LST-Net.

Robot arms should be able to learn new tasks. One framework here is reinforcement learning, where the robot is given a reward function that encodes the task, and the robot autonomously learns actions to maximize its reward. Existing approaches to reinforcement learning often frame this problem as a Markov decision process, and learn a policy (or a hierarchy of policies) to complete the task. These policies reason over hundreds of fine-grained actions that the robot arm needs to take: e.g., moving slightly to the right or rotating the end-effector a few degrees. But the manipulation tasks that we want robots to perform can often be broken down into a small number of high-level motions: e.g., reaching an object or turning a handle. In this paper we therefore propose a waypoint-based approach for model-free reinforcement learning. Instead of learning a low-level policy, the robot now learns a trajectory of waypoints, and then interpolates between those waypoints using existing controllers. Our key novelty is framing this waypoint-based setting as a sequence of multi-armed bandits: each bandit problem corresponds to one waypoint along the robot’s motion. We theoretically show that an ideal solution to this reformulation has lower regret bounds than standard frameworks. We also introduce an approximate posterior sampling solution that builds the robot’s motion one waypoint at a time. Results across benchmark simulations and two real-world experiments suggest that this proposed approach learns new tasks more quickly than state-of-the-art baselines. See our website here: https://collab.me.vt.edu/rl-waypoints/

Robotic waste sorting poses significant challenges in both perception and manipulation, given the extreme variability of objects that should be recognized on a cluttered conveyor belt. While deep learning has proven effective in solving complex tasks, the necessity for extensive data collection and labeling limits its applicability in real-world scenarios like waste sorting. To tackle this issue, we introduce a data augmentation method based on a novel GAN architecture called wasteGAN. The proposed method allows to increase the performance of semantic segmentation models, starting from a very limited bunch of labeled examples, such as few as 100. The key innovations of wasteGAN include a novel loss function, a novel activation function, and a larger generator block. Overall, such innovations helps the network to learn from limited number of examples and synthesize data that better mirrors real-world distributions. We then leverage the higher-quality segmentation masks predicted from models trained on the wasteGAN synthetic data to compute semantic-aware grasp poses, enabling a robotic arm to effectively recognizing contaminants and separating waste in a real-world scenario. Through comprehensive evaluation encompassing dataset-based assessments and real-world experiments, our methodology demonstrated promising potential for robotic waste sorting, yielding performance gains of up to 5.8% in picking contaminants. The project page is available at https://github.com/bach05/wasteGAN.git.