ICRA 2023 Papers

Fast and safe manipulation of flexible objects with a robot manipulator necessitates measures to cope with vibrations. Existing approaches either increase the task execution time or require complex models and/or additional instrumentation to measure vibrations. This paper develops a model-based method that overcomes these limitations. It relies on a simple pendulum-like model for modeling the beam, open-loop optimal control for suppressing vibrations, and does not require any exteroceptive sensors. We experimentally show that the proposed method drastically reduces residual vibrations – at least 90% – and outperforms the commonly used input shaping (IS) for trajectories with the same execution time. Besides, our method can also execute the task faster than IS with a minor reduction in vibration suppression performance, thereby facilitating the development of new solutions for flexible object manipulation tasks.

Autonomous mobile robots (AMRs) have the capability to execute a wide range of tasks with minimal human intervention. However, one of the major limitations of AMRs is their limited battery life, which often results in interruptions to their task execution and the need to reach the nearest charging station. Optimizing energy consumption in AMRs has become a critical challenge in their deployment. Through empirical studies on real AMRs, we have identified a lack of coordination between computation and control as a major source of energy inefficiency. In this paper, we propose a comprehensive energy prediction model that provides real-time energy consumption for each component of the AMR. Additionally, we propose three path models to address the obstacle avoidance problem for AMRs. To evaluate the performance of our energy prediction and path models, we have developed a customized AMR called Donkey, which has the capability for fine-grained (millisecond-level) end-to-end power profiling. Our energy prediction model demonstrated an accuracy of over 90% in our evaluations. Finally, we applied our energy prediction model to obstacle avoidance and guided energy-efficient path selection, resulting in up to a 44.8% reduction in energy consumption compared to the baseline.

We present a generalised architecture for reactive mobile manipulation while a robot's base is in motion toward the next objective in a high-level task. By performing tasks on-the-move, overall cycle time is reduced compared to methods where the base pauses during manipulation. Reactive control of the manipulator enables grasping objects with unpredictable motion while improving robustness against perception errors, environmental disturbances, and inaccurate robot control compared to open-loop, trajectory-based planning approaches. We present an example implementation of the architecture and investigate the performance on a series of pick and place tasks with both static and dynamic objects and compare the performance to baseline methods. Our method demonstrated a real-world success rate of over 99%, failing in only a single trial from 120 attempts with a physical robot system. The architecture is further demonstrated on other mobile manipulator platforms in simulation. Our approach reduces task time by up to 48%, while also improving reliability, gracefulness, and predictability compared to existing architectures for mobile manipulation. See benburgesslimerick.github.io/ManipulationOnTheMove for supplementary materials.

Robot manipulation today generally focuses on motions exclusively with a robot arm or a dexterous hand, but usually not a combination of both. However, complex manipulation tasks can require coordinating arm and hand motions that leverage capabilities of both, much like the coordinated arm and hand motions carried out by humans to perform everyday tasks. In this work, we evaluate unified manipulation with robot arms and dexterous hands, using a motion optimization framework that synthesizes a series of configuration states over the entire manipulation system. We characterize the possible benefits of unifying arm and dexterous hand capabilities within a single model via metrics such as pose accuracy, manipulability, joint-space smoothness, distance to joint-limits, distance to collisions, and more. Several arm-hand combinations are quantitatively compared in simulation on a variety of experiment tasks and performance measures. Our results suggest that combining motions from robot arms and dexterous hands indeed has compelling benefits, highlighting the exciting potential of continued progress in unified arm-hand motion synthesis for robotics applications.

Future NASA lander missions to icy moons will require completely automated, accurate, and data efficient calibration methods for the robot manipulator arms that sample icy terrains in the lander's vicinity. To support this need, this paper presents a Gaussian Process (GP) approach to the classical manipulator kinematic calibration process. Instead of identifying a corrected set of Denavit-Hartenberg kinematic parameters, a set of GPs models the residual kinematic error of the arm over the workspace. More importantly, this modeling framework allows a Gaussian Process Upper Confident Bound (GP-UCB) algorithm to efficiently and adaptively select the calibration's measurement points so as to minimize the number of experiments, and therefore minimize the time needed for recalibration. The method is demonstrated in simulation on a simple 2-DOF arm, a 6 DOF arm whose geometry is a candidate for a future NASA mission, and a 7 DOF Barrett WAM arm.

In recent years, impressive results have been achieved in robotic manipulation. While many efforts focus on generating collision-free reference signals, few allow safe contact between the robot bodies and the environment. However, in human's daily manipulation, contact between arms and obstacles is prevalent and even necessary. This paper investigates the benefit of allowing safe contact during robotic manipulation and advocates generating and tracking compliance reference signals in both operational and null spaces. In addition, to optimize the collision-allowed trajectories, we present a hybrid solver that integrates sampling- and gradient-based approaches. We evaluate the proposed method on a goal-reaching task in five simulated and real-world environments with different collisional conditions. We show that allowing safe contact improves goal-reaching efficiency and provides feasible solutions in highly collisional scenarios where collision-free constraints cannot be enforced. Moreover, we demonstrate that planning in null space, in addition to operational space, improves trajectory safety. Further information is available at https://rolandzhu.github.io/ContactReach/.

Practical implementations of deep reinforcement learning (deep RL) have been challenging due to an amplitude of factors, such as designing reward functions that cover every possible interaction. To address the heavy burden of robot reward engineering, we aim to leverage subjective human preferences gathered in the context of human-robot interaction, while taking advantage of a baseline reward function when available. By considering baseline objectives to be designed beforehand, we are able to narrow down the policy space, solely requesting human attention when their input matters the most. To allow for control over the optimization of different objectives, our approach contemplates a multi-objective setting. We achieve human-compliant policies by sequentially training an optimal policy from a baseline specification and collecting queries on pairs of trajectories. These policies are obtained by training a reward estimator to generate Pareto optimal policies that include human preferred behaviours. Our approach ensures sample efficiency and we conducted a user study to collect real human preferences, which we utilized to obtain a policy on a social navigation environment.

Detect-and-Avoid (DAA) capabilities are critical for safe operations of unmanned aircraft systems (UAS). This paper introduces, AirTrack, a real-time vision-only detect and tracking framework that respects the size, weight, and power (SWaP) constraints of sUAS systems. Given the low Signal-to-Noise ratios (SNR) of far away aircraft, we propose using full resolution images in a deep learning framework that aligns successive images to remove ego-motion. The aligned images are then used downstream in cascaded primary and secondary classifiers to improve detection and tracking performance on multiple metrics. We show that AirTrack outperforms state-of-the art baselines on the Amazon Airborne Object Tracking (AOT) Dataset. Multiple real world flight tests with a Cessna 182 interacting with general aviation traffic and additional near-collision flight tests with a Bell helicopter flying towards a UAS in a controlled setting showcase that the proposed approach satisfies the newly introduced ASTM F3442/F3442M standard for DAA. Empirical evaluations show that our system has a probability of track of more than 95% up to a range of 700m. [Video]11Video: https://youtu.be/bMw5nUGL5GQ

This work developed a kernel-based residual learning framework for quadrupedal robotic locomotion. Ini-tially, a kernel neural network is trained with data collected from an MPC controller. Alongside a frozen kernel network, a residual controller network is trained using reinforcement learning to acquire generalized locomotion skills and robust-ness against external perturbations. The proposed framework successfully learns a robust quadrupedal locomotion controller with high sample efficiency and controllability, which can provide omnidirectional locomotion at continuous velocities. We validated its versatility and robustness on unseen terrains that the expert MPC controller failed to traverse. Furthermore, the learned kernel can produce a range of functional locomotion behaviors and can generalize to unseen gaits.

Real-time navigation in non-trivial environments by micro aerial vehicles (MAVs) predominantly relies on modelling the MAV with idealized geometry, such as a sphere. Simplified, conservative representations increase the likelihood of a planner failing to identify valid paths. That likelihood increases the more a robot's geometry differs from the idealized version. Few current approaches consider these situations; we are unaware of any that do so using perception space representations. This work introduces the egocan, a perception space obstacle representation using line-of-sight free space estimates, and 3D Gap, a perception space approach to gap finding for identifying goal-directed, collision-free directions of travel through 3D space. Both are integrated, with real-time considerations in mind, to define a local planner module of a hierarchical navigation system. The result is Aerial Local Planning in Perception Space (AeriaLPiPS). AeriaLPiPS is shown to be capable of safely navigating a MAV with non-idealized geometry through various environments, including those impassable by traditional real-time approaches. The open source implementation of this work is available at github.com/ivaROS/AeriaLPiPS.

Reinforcement learning (RL) has emerged as a powerful approach for locomotion control of highly articulated robotic systems. However, one major challenge is the tedious process of tuning the reward function to achieve the desired motion style. To address this issue, imitation learning approaches such as adversarial motion priors have been proposed, which encourage a pre-defined motion style. In this work, we present an approach to enhance the concept of adversarial motion prior-based RL, allowing for multiple, discretely switchable motion styles. Our approach demonstrates that multiple styles and skills can be learned simultaneously without significant performance differences, even in combination with motion data-free skills. We conducted several real-world experiments using a wheeled-legged robot to validate our approach. The experiments involved learning skills from existing RL controllers and trajectory optimization, such as ducking and walking, as well as novel skills, such as switching between a quadrupedal and humanoid configuration. For the latter skill, the robot was required to stand up, navigate on two wheels, and sit down. Instead of manually tuning the sit-down motion, we found that a reverse playback of the stand-up movement helped the robot discover feasible sit-down behaviors and avoided the need for tedious reward function tuning.

Event-based cameras have recently shown great potential for high-speed motion estimation owing to their ability to capture temporally rich information asynchronously. Spiking Neural Networks (SNNs), with their neuro-inspired event-driven processing can efficiently handle such asynchronous data, while neuron models such as the leaky-integrate and fire (LIF) can keep track of the quintessential timing information contained in the inputs. SNNs achieve this by maintaining a dynamic state in the neuron memory, retaining important information while forgetting redundant data over time. Thus, we posit that SNNs would allow for better performance on sequential regression tasks compared to similarly sized Analog Neural Networks (ANNs). However, deep SNNs are difficult to train due to vanishing spikes at later layers. To that effect, we propose an adaptive fully-spiking framework with learnable neuronal dynamics to alleviate the spike vanishing problem. We utilize surrogate gradient-based backpropagation through time (BPTT) to train our deep SNNs from scratch. We validate our approach for the task of optical flow estimation on the Multi-Vehicle Stereo Event-Camera (MVSEC) dataset and the DSEC-Flow dataset. Our experiments on these datasets show an average reduction of ∼ 13% in average endpoint error (AEE) compared to state-of-the-art ANNs. We also explore several down-scaled models and observe that our SNN models consistently outperform similarly sized ANNs offering ∼10%-16% lower AEE. These results demonstrate the importance of SNNs for smaller models and their suitability at the edge. In terms of efficiency, our SNNs offer substantial savings in network parameters (∼ 48.3 ×) and computational energy (∼ 10.2 ×) while attaining ∼ 10% lower EPE compared to the state-of-the-art ANN implementations.

Stroke affects a large number of people every year. One consequence is the weakness of ambulatory muscles resulting in a paretic gait. Actuated ankle foot orthoses can be a solution to assist paretic patients to dorsiflex and/or plantar flex their ankle joint during the gait phases. To assist the wearer following a predefined ankle joint desired trajectory, an adaptive active disturbance rejection controller is proposed in this study. The human muscular torque and estimation errors are estimated through a nonlinear disturbance observer based on the estimated model. This estimated torque is compensated within the proposed projection based adaptive controller combined to a saturated proportional derivative term. The purposes of using this controller are: i) the no need of prior system's parameter identification due to the adaptive structure, ii) the assistance-as-needed of the wearer through the rejection term and iii) the avoidance of the actuator saturation by including projection and saturation functions. This controller is tested in real time using an actuated ankle-foot-orthosis (AAFO) in lab environment with three healthy subjects to show its effectiveness.

Trajectory optimisation methods for robotic motion planning often require the use of first order derivatives of the dynamics of the system with respect to the states and controls of the system. Particularly when multi-contact dynamics are present, these derivatives are often numerically approximated by a method such as finite-differencing. Finite-differencing whilst using an expensive physics simulator is usually the bottleneck in these trajectory optimisation algorithms. Since these dynamics derivatives do not change substantially over certain time inter-vals, we propose that trajectory optimisers can compute the dy-namics derivatives less often and then interpolate approximations to the derivatives in between calculated derivatives, gaining a sig-nificant speed up for overall optimisation time with no observable degradation in the generated behaviour. We investigate different methods of interpolating approximations as well as propose an adaptive method to detect when to compute the derivatives with finite-differencing. We find a speed-up of planning times on average by 60% in a contact-based manipulation task.

For robotic vehicles to navigate robustly and safely in unseen environments, it is crucial to decide the most suitable navigation policy. However, most existing deep reinforcement learning based navigation policies are trained with a hand-engineered curriculum and reward function which are difficult to be deployed in a wide range of real-world scenarios. In this paper, we propose a framework to learn a family of low-level navigation policies and a high-level policy for deploying them. The main idea is that, instead of learning a single navigation policy with a fixed reward function, we simultaneously learn a family of policies that exhibit different behaviors with a wide range of reward functions. We then train the high-level policy which adaptively deploys the most suitable navigation skill. We evaluate our approach in simulation and the real world and demonstrate that our method can learn diverse navigation skills and adaptively deploy them. We also illustrate that our proposed hierarchical learning framework presents explainability by providing semantics for the behavior of an autonomous agent.

In this paper, we present a Computer Vision (CV) based tracking and fusion algorithm, dedicated to a 3D printed gimbal system on drones flying in nature. The whole gimbal system can stabilize the camera orientation robustly in challenging environments by using skyline and ground plane as references. Our main contributions are the following: a) a light-weight Resnet-18 backbone network model was trained from scratch, and deployed onto the Jetson Nano platform to segment the image specifically into binary parts (ground and sky); b) our geometry assumption from the skyline and ground cues delivers the potential for robust visual tracking in the wild by using the skyline and ground plane as references; c) a manifold surface-based adaptive particle sampling can fuse orientation from multiple sensor sources flexibly. The whole algorithm pipeline is tested on our 3D-printed gimbal module with Jetson Nano. The experiments were performed on top of a building in a real landscape. The public code link: https://github.com/alexandor91/gimbal-fusion.git.

Enabling the capability of assessing risk and making risk-aware decisions is essential to applying reinforcement learning to safety-critical robots like drones. In this paper, we investigate a specific case where a nano quadcopter robot learns to navigate an apriori-unknown cluttered environment under partial observability. We present a distributional reinforcement learning framework to generate adaptive risk-tendency policies. Specifically, we propose to use lower tail conditional variance of the learnt return distribution as intrinsic uncertainty estimation, and use exponentially weighted average forecasting (EWAF) to adapt the risk-tendency in accordance with the estimated uncertainty. In simulation and real-world empirical results, we show that (1) the most effective risk-tendency varies across states, (2) the agent with adaptive risk-tendency achieves superior performance compared to risk-neutral policy or risk-averse policy baselines. Code and video can be found in this repository: https://github.com/tudelft/risk-sensitive-rl.git

It is critical to perceive physical contact for intelligent robots to safely interact in dynamic, unstructured environments. As physical contacts can occur at any location, a well-performing tactile sensing system should be able to deploy a large area on robotic surface. Some researchers have implemented large-area tactile sensors by using sensing arrays, but it is challenging to deploy many sensing elements. Electrical resistance tomography (ERT) has recently been introduced into tactile sensing to overcome some of the limitations with conventional tactile sensing arrays, and good results have been achieved for some robotic applications. However, a particular challenge is that spatial resolution is low. Although various attempts have been made to improve the performance of ERT-based tactile sensors, the intrinsic resolution issue remains unsolved. In this paper, we propose a novel adaptive optimal drive strategy for efficient ERT-based large-area tactile sensing for robotic applications, which can adaptively select the current injection and voltage measurement pattern for optimal tactile stimulus. In particular, regions of tactile contacts are preliminarily detected and localized by a base scanning pattern with only a few measurement data. According to this detected region, the adaptive strategy can select the optimal current injection and voltage measurement pattern to improve the sensing performance by maximizing the current density. To verify the effectiveness of the proposed strategy, the proposed method is comprehensively evaluated by simulation and experiments. The results revealed that the optimal strategy can effectively improve both spatial and temporal resolution.

In this paper, we present a LiDAR Inertial Odometry (LIO) algorithm utilizing adaptive keyframe generation which achieves fast and accurate state estimation for aerial and ground robots. It is known that keyframe generation significantly affects the performance of Simultaneous Localization and Mapping (SLAM) algorithms. Unlike existing SLAM algorithms that generate keyframes based on fixed conditions, we propose to use adaptive keyframe generation conditions considering characteristics of surrounding environment using real-time LiDAR scans. When a keyframe is generated, the keyframe and the corresponding LiDAR measurements are stored in our novel data structure designed for efficient sub- map generation. The scan to sub-map matching module then uses the Generalized Iterative Closest Point (GICP) algorithm to adjust estimated states at a global scale, producing more accurate and globally consistent state estimation results even in large-scale underground environments. Experimental results from diverse types of underground environments show that the proposed method outperforms the existing state-of-the-art LIO algorithms in various metrics such as computational speed, CPU usage, and accuracy.