IROS 2022 Papers

Learned knowledge graph representations supporting robots contain a wealth of domain knowledge that drives robot behavior. However, there does not exist an inference reconciliation framework that expresses how a knowledge graph representation affects a robot's sequential decision making. We use a pedagogical approach to explain the inferences of a learned, black-box knowledge graph representation, a knowledge graph embedding. Our interpretable model uses a decision tree classifier to locally approximate the predictions of the black-box model and provides natural language explanations interpretable by non-experts. Results from our algorithmic evaluation affirm our model design choices, and the results of our user studies with non-experts support the need for the proposed inference reconciliation framework. Critically, results from our simulated robot evaluation indicate that our explanations enable non-experts to correct erratic robot behaviors due to nonsensical beliefs within the black-box.

Industrial robots require force controllability and impact resistance to ensure safe physical interactions. An electro-hydrostaic actuator (EHA) is expected to be suitable for such applications because it has high backdrivability which improve both force controllability at contact and impact resistance. However, EHAs had been rarely used in multi-axes robotic systems. The previous works validated the force controllability of the EHA-driven robot Hydra. However, the impact resistance of an EHA-driven robot is still unclear. In order to evaluate the impact resistance of the high-power EHA-driven robot, we developed high-pressure EHAs employing ceramics as rigid material to reduce internal leakage, and developed the EHA-driven 6-DOF robot arm Hydracer as the platform for the evaluation. This paper describes the mechanism of Hydracer especially on the base 3-DOF mechanism, and conducts the backdrivability measurement and the impact resistance evaluation.

This paper demonstrates the practical performance of a new theory of balance control that has been shown in simulation to out-perform earlier balance control theories in the sense of allowing the robot to make larger and faster movements while still maintaining its balance. The case studied here is that of a general planar double inverted pendulum, which resembles a legged robot's behaviour when the polygon of support shrinks to a line. The results show the speed and accuracy of the controller, as well as its robustness to external disturbances and slipping during fast movements.

Lower-limb exoskeletons are robotic devices that can provide assistance to human locomotion. Since they are expected to be used in ecological environments, their control strategy should handle different kinds of daily-life situations. Taking inspiration from the human neuromuscular system - and particularly from the socalled motor primitives - may help in adapting the type of delivered assistance to different locomotion tasks. In this work, we validated the combination of simplified primitives and a musculoskeletal model for assisting healthy subjects with a hip exoskeleton. This framework showed adaptation to the user's gait for different slope inclinations, although its effects on the subject's speed and their perceived effort showed no significant improvement compared to wearing the device in transparent mode.

This paper presents a multi-modal model-based reinforcement learning (MBRL) approach to the excavation of fragmented rocks, which are very challenging to model due to their highly variable sizes and geometries, and visual occlusions. A multi-modal recurrent neural network (RNN) learns the dynamics of bucket-terrain interaction from a small physical dataset, with a discrete set of motion primitives encoded with domain knowledge as the action space. Then a model predictive controller (MPC) tracks a global reference path using multi-modal feedback. We show that our RNN-based dynamics function achieves lower prediction errors compared to a feed-forward neural network baseline, and the MPC is able to significantly outperform manually designed strategies on such a challenging task.

Unmanned aerial vehicles (UAVs) are becoming more common, presenting the need for effective human-robot communication strategies that address the unique nature of unmanned aerial flight. Visual communication via drone flight paths, also called gestures, may prove to be an ideal method. However, the effectiveness of visual communication techniques is dependent on several factors including an observer's position relative to a UAV. Previous work has studied the maximum line-of-sight at which observers can identify a small UAV [1]. However, this work did not consider how changes in distance may affect an observer's ability to perceive the shape of a UAV's motion. In this study, we conduct a series of online surveys to evaluate how changes in line-of-sight distance and gesture size affect observers' ability to identify and distinguish between UAV gestures. We first examine observers' ability to accurately identify gestures when adjusting a gesture's size relative to the size of a UAV. We then measure how observers' ability to identify gestures changes with respect to varying line-of-sight distances. Lastly, we consider how altering the size of a UAV gesture may improve an observer's ability to identify drone gestures from varying distances. Our results show that increasing the gesture size across varying UAV to gesture ratios did not have a significant effect on participant response accuracy. We found that between 17 m and 75 m from the observer, their ability to accurately identify a drone gesture was inversely proportional to the distance between the observer and the drone. Finally, we found that maintaining a gesture's apparent size improves participant response accuracy over changing line-of-sight distances.

Extensive research has been conducted on autonomous surface robots and underwater robots for various tasks in aquatic environments. The duration of the operation of autonomous field robots depends on the capacity of the mounted battery, as they are not typically connected to an external power supply. Therefore, smart strategies which are optimized for each task are required to extend the working time of autonomous field robots. We have developed a symmetrically-shaped au-tonomous surface robot for the long-term monitoring of water quality. In this study, we propose position-keeping strategies to prolong the duration of the symmetrically-shaped surface robot for in-situ monitoring. The proposed position-keeping strategies are evaluated in terms of the power consumption and mean error distance in both practical and simulation environments. The experimental results demonstrate that a robot placed on a water surface with disturbance determines the best course of action to maintain its position based on the environmental conditions and application.

A robot must comply with very restrictive safety standards in close human-robot collaboration applications. These standards limit the robot's performance because of speed reductions to avoid potentially large forces exerted on humans during collisions. On-robot capacitive proximity sensors (CPS) can serve as a solution to allow higher speeds and thus better productivity. They allow early reactive measures before contacts occur to reduce the forces during collisions. An open question on designing the systems is the selection of an adequate activation distance to trigger safety measures for a specific robot while considering latency and detection robustness. Furthermore, the systems' actual effectiveness of impact attenuation and performance gain has not been evaluated before. In this work, we define and conduct a unified test procedure based on collision experiments to determine these parameters and investigate the performance gain. Two capacitive proximity sensor systems are evaluated on this test strategy on two robots. A significant performance increase can be achieved, since a small detection distance doubles robot operation speed while maintaining the same contact force as without Capacitive Proximity Sensor (CPS). This work can serve as a reference guide for designing, configuring and implementing future on-robot CPS.

We seek to rigorously evaluate the benefit of using a few beams rather than a single beam for a low-cost obstacle avoidance sonar for small AUVs. For a small low-cost AUV, the complexity, cost, and volume required for a multi-beam forward looking sonar are prohibitive. In contrast, a single-beam system is relatively easy to integrate into a small AUV, but does not provide the performance of a multi-beam solution. To better understand this trade-off, we seek to rigorously quantify the improvement with respect to obstacle avoidance performance of adding just a few beams to a single-beam forward looking sonar relative to the performance of the single-beam system. Our work fundamentally supports the goal of using small low-cost AUV systems in cluttered and unstructured environments. Specifically, we investigate the benefit of incorporating a port and starboard beam to a single-beam sonar system for collision avoidance. A methodology for collision avoidance is developed to obtain a fair comparison between a single-beam and multi-beam system, explicitly incorporating the geometry of the beam patterns from forward-looking sonars with large beam angles, and simulated using a high-fidelity representation of acoustic signal propagation.

Explainable artificial intelligence is a research field that tries to provide more transparency for autonomous intelligent systems. Explainability has been used, particularly in reinforcement learning and robotic scenarios, to better understand the robot decision-making process. Previous work, however, has been widely focused on providing technical explanations that can be better understood by AI practitioners than non-expert end-users. In this work, we make use of human-like explanations built from the probability of success to complete the goal that an autonomous robot shows after performing an action. These explanations are intended to be understood by people who have no or very little experience with artificial intelligence methods. This paper presents a user trial to study whether these explanations that focus on the probability an action has of succeeding in its goal constitute a suitable explanation for non-expert end-users. The results obtained show that non-expert participants rate robot explanations that focus on the probability of success higher and with less variance than technical explanations generated from Q-values, and also favor counterfactual explanations over standalone explanations.

This research focuses on soft robotic bladders that are used to monitor and control the interaction between a user's head and the shell of a Smart Helmet. Compression of these bladders determines impact dissipation; hence the focus of this paper is sensing and estimation of bladder compression. An IR rangefinder-based solution is evaluated using regression techniques as well as a Neural Network to estimate bladder compression. A Hall-Effect (HE) magnetic sensing system is also examined where HE sensors embedded in the base of the bladder sense the position of a magnet in the top of the bladder. The paper presents the HE sensor array, signal processing of HE voltage data, and then a Neural Network (NN) for predicting bladder compression. Efficacy of different training data sets on NN performance is studied. Different NN configurations are examined to determine a configuration that provides accurate estimates with as few nodes as possible. Different bladder compression profiles are evaluated to characterize IR range finding and HE based techniques in application scenarios.

Exploring unknown environments, such as caves or planetary surfaces, requires a quick understanding of the surroundings. Beforehand, only aerial footage from satellites or images from previous missions might be available. The proposed ensemble based anomaly detection framework utilizes previously gained knowledge and incorporates it with insights gained during the mission. The modular system consists of different networks which are combined to determine anomalies in the current surroundings. By utilizing data from other missions, simulations or aerial photos, a precise anomaly detection can be achieved at the start of a mission. The system can further be improved by training new networks during the mission, which can be incorporated into the ensemble at runtime. This allows for synchronous execution of mission and training of models on a base station. The proposed system is tested and evaluated on an ANYmal C walking robot in different scenarios, however the approach is applicable for different kinds of mobile robots. The results show a clear improvement of ensembles compared to individual networks, while keeping a small memory footprint and low inference time on the mobile system.

In nature, the movement of quadrupeds is completed under the combined action of the spine and the legs. Inspired by this, this paper explores the effect of a lateral flexing spine on the locomotion of a rat robot. Benefiting from the regular lateral flexion of a soft actuated spine, the rat robot exhibits enhance step length of its hind legs and increased translational velocity by coordinating the opposite movements of the left and right sides. Furthermore, this paper introduces a mathematical model of the effect of the flexible spine on the robot velocity. Finally, extensive experiments are conducted in simulations and on the physical rat robot. Compared with the locomotion without a flexing spine, the simulation results show that the velocity of the robot can be increased up to 218.29%, which is in line with the theoretical results from the proposed mathematical model. Limited by the gap between simulation and the real world, the experiment results of the physical rat robot show a slight performance than the theoretical results. But the physical rat robot can still enhance its translational velocity with the help of a lateral flexing spine.

This work introduces a Nonlinear Model Predictive Control (N-MPC) for camera-equipped Unmanned Aerial Vehicles (UAVs), which controls at the motor level the UAV motion to ensure the quality of vision-based state estimation while performing other tasks. The controller ensures visibility over a sufficient amount of features, while optimizing their coverage, based on an assessment of the estimation quality. The controller works for the very broad class of generic multirotor UAVs, including platforms with any number of propellers, which can be both collinear, as in the quadrotor, and fixedly-tilted. The low-level inputs are computed in real-time and realistically constrained, in terms of maximum motor torque. This allows the platform to exploit its full actuation capabilities to maintain the visibility over the set of points of interest. Our implementation is tested in Gazebo simulations and in mocap-free real experiments, and features a visual-inertial state estimation based on Kalman filter. The software is provided open-source.

The optimal stiffness for soft swimming robots depends on swimming speed, which means no single stiffness can maximise efficiency in all swimming conditions. Tunable-stiffness would produce an increased range of high-efficiency swimming speeds for robots with flexible propulsors and enable soft control surfaces for steering underwater vehicles. We propose and demonstrate a method for tunable soft robotic stiffness using inflatable rubber tubes to stiffen a silicone foil through pressure and second moment of area change. We achieved double the effective stiffness of the system for an input pressure change from 0 to 0.8 bar and 2 J energy input. We achieved a resonant amplitude gain of 5 to 7 times the input amplitude and tripled the high-gain frequency range compared to a foil with fixed stiffness. These results show that changing second moment of area is an energy effective approach to tunable-stiffness robots.

In many environmental monitoring applications robots are often tasked to visit various distinct locations to make observations and/or collect specific measurements. The problem of scheduling and assigning robots to the various tasks and planning feasible paths for the robots can be posed as an Orienteering Problem (OP). In the standard OP, routing and scheduling is achieved by maximizing an objective function by visiting the most rewarding locations while respecting a limited travel budget. However, traditional formulations for such problems usually neglect some environmental features that can greatly impact the tour, e.g., flows, such as wind or ocean currents. This is of particular importance for applications in marine and atmospheric environments where vehicle motions can be significantly impacted by the environmental dynamics and the environment exerts a non-negligible force on the vehicles. In this paper, we tackle the OP in fluid environments where robots must operate in the presence of ocean and/or atmospheric currents. We introduce a novel multi-objective formulation that combines both task and path planning problems, and whose goals are to (i) maximize the collected reward, while (ii) minimizing the energy expenditure by leveraging the environmental dynamics wherever possible. We validate our strategy using simulated ocean model data to show that our approach can generate a diverse set of solutions that have an adequate compromise between both objectives.

In this paper, we present an online planning-scheduling approach for battery-powered autonomous aerial robots. The approach consists of simultaneously planning a coverage path and scheduling onboard computational tasks. We further derive a novel variable coverage motion robust to air-borne constraints and an empirically motivated energy model. The model includes the energy contribution of the schedule based on an automatic computational energy modeling tool. Our experiments show how an initial flight plan is adjusted online as a function of the available battery, accounting for uncertainty. Our approach remedies possible in-flight failure in case of unexpected battery drops, e.g., due to adverse atmospheric conditions, and increases the overall fault tolerance.

Object feature pollution is one of the burning issues in vision-based UAV tracking, commonly caused by occlusion, fast motion, and illumination variation. Due to the contaminated information in the polluted object features, most trackers fail to precisely estimate the object location and scale. To address the above disturbing issue, this work proposes a novel end-to-end feature decontaminated network for efficient and effective UAV tracking, i.e., FDNT. FDNT mainly includes two modules: a decontaminated downsampling network and a decontaminated upsampling network. The former reduces the interference information of the feature pollution and enhanced the expression of the object location information with two asymmetric convolution branches. The latter restores the object scale information with the super-resolution technology-based low-to-high encoder, achieving a further decontamination effect. Moreover, a novel pooling distance loss is carefully developed to assist the decontaminated downsampling network in concentrating on the critical regions with the object information. Exhaustive experiments on three well-known benchmarks validate the effectiveness of FDNT, especially on the sequences with feature pollution. In addition, real-world tests show the efficiency of FDNT with 31.4 frames per second. The code and demo videos are available at https://github.com/vision4robotics/FDNT.

This study focuses on replicating the muscu-loskeletal system of human arms for mimicking its movement. Muscle redundancy is critical for regulating the mechanical impedance of arms and legs. However, when implementing muscle redundancy on robots, making an ill-posed problem that cannot determine the muscle forces uniquely. In this paper, first, a method for controlling end-point stiffness in the muscle space for the joint and muscle redundant system is described. Next, the muscle model imitating the nonlinear viscosity characteristic of human muscles is introduced. Then, a method to control the joint viscosity by adjusting the internal forces of muscles adequately without affecting the stiffness control directly is proposed. Finally, numerical simulations are performed to investigate the effectiveness of the proposed method.