IROS 2020 Papers

Human-Robot Interaction (HRI) user studies are challenging to evaluate and compare due to a lack of standardization and the infrastructure required to implement each study. The lack of experimental infrastructure also makes it difficult to systematically evaluate the impact of individual components (e.g., the quality of perception software) on overall system performance. This work proposes a framework to ease the implementation and reproducibility of human-robot interaction user studies. The framework utilizes ROS middleware and is implemented with four modules: perception, decision, action, and metrics. The perception module aggregates sensor data to be used by the decision and action modules. The decision module is the task-level executive and can be designed by the HRI researcher for their specific task. The action module takes subtask requests from the decision module and breaks them down into motion primitives for execution on the robot. The metrics module tracks and generates quantitative metrics for the study. The framework is implemented with modular interfaces to allow for alternate implementations within each module and can be generalized for a variety of tasks and human/robot roles. The framework is illustrated through an example scenario involving a human and a Franka Emika Panda arm collaboratively assembling a toolbox together.

Having the merits of chemical inertness and immunity to electromagnetic interference, light weight, small size, and softness, optical waveguides have attracted much attention in making tactile sensors recently. This paper presents a new design of waveguide using two layers of cores, one of which has an uniform width and the other has an incremental width. It is deduced and verified that the contact force can be derived from the light power loss in the uniform-width core, while the contact position can be derived from the light power loss in the other core together with the estimated force. By this dual-core design, a single waveguide can simultaneously and continuously measure the contact force and position along it, which makes it very suited for integration on some thin long robotic parts, such as robotic fingers. A hardware experiment has been conducted to demonstrate its effectiveness on a two-finger gripper in an assembly task. The dual-core waveguide achieves 2 mm spatial resolution and 0.1 N sensitivity.

Place recognition is a core component of Simultaneous Localization and Mapping (SLAM) algorithms. Particularly in visual SLAM systems, previously-visited places are recognized by measuring the appearance similarity between images representing these locations. However, such approaches are sensitive to visual appearance change and also can be computationally expensive. In this paper, we propose an alternative approach adapting LiDAR descriptors for 3D points obtained from stereo-visual odometry for place recognition. 3D points are potentially more reliable than 2D visual cues (e.g., 2D features) against environmental changes (e.g., variable illumination) and this may benefit visual SLAM systems in long-term deployment scenarios. Stereo-visual odometry generates 3D points with an absolute scale, which enables us to use LiDAR descriptors for place recognition with high computational efficiency. Through extensive evaluations on standard benchmark datasets, we demonstrate the accuracy, efficiency, and robustness of using 3D points for place recognition over 2D methods.

This paper proposes a distributed field mapping algorithm that drives a team of robots to explore and learn an unknown scalar field. The algorithm is based on a bio-inspired approach known as Speeding-Up and Slowing-Down (SUSD) for distributed source seeking problems. Our algorithm leverages a Gaussian Process model to predict field values as robots explore. By comparing Gaussian Process predictions with measurements of the field, agents search along the gradient of the model error while simultaneously improving the Gaussian Process model. We provide a proof of convergence to the gradient direction and demonstrate our approach in simulation and experiments using 2D wheeled robots and 2D flying autonomous miniature blimps.

The challenges of developing low-cost, large-scale multi-robot navigation systems include noisy measurements, a large number of robots, and computing efficiency for collision avoidance. This paper presents a distributed motion planning framework for a large number of robots to navigate with robust collision avoidance using low-cost range only measurements. The novelty of this work is threefold. (1) Developing a distributed collision-free navigation system for a large-scale robot group in which each robot performs motion planning based on the noisy range measurements of neighboring robots; (2) Developing a set of algorithms for each robot to accurately estimate the relative positions and orientations based on the range measurements and relative velocities; (3) Developing a velocity obstacle (VO) based motion planning algorithm for each robot which can take into account of the estimation uncertainties in relative positions and orientations. The proposed approach is tested with various numbers of differential-driven robots in the Gazebo simulator and real-world experiments. Both simulation and experiment results validate the superior performance of the proposed approach compared to other state-of-art technologies.

To function effectively in real-world environments, powered wearable robots such as exoskeletons and robotic prostheses must recognize the user's motion intent by detecting the user's locomotion modes such as walking, stair ascent and descent or ramp ascent and descent. Traditionally, intent detection is achieved using rule based methods such as state machines or fuzzy logic using data from wearable sensors. Due to the difficulty of manual rule design, these methods are limited to detect certain simple locomotion modes. Machine learning (ML) based methods can perform classification on a large number of classes without manual rule design and recent research has explored several ML methods for locomotion mode classification. However, current ML based methods for locomotion mode detection use classical methods that require use of feature engineering to achieve acceptable accuracies. Additionally, current ML strategies only classify when certain motion events are detected. This strategy, while computationally efficient could result in misclassifications affecting large sections of motion recognition. To overcome these limitations, this paper proposes an end-to-end deep learning based method for locomotion mode detection that eliminates the need for feature engineering and classifies at a fixed sample rate. This paper introduces a new metric called confidence index and proposes a strategy for tuning confidence index thresholds to achieve a stable intent recognition and overall accuracy of greater than 95% on a publicly available benchmark dataset.

For safe and efficient human-robot interaction, a robot needs to predict and understand the intentions of humans who share the same space. Mobile robots are traditionally built to be reactive, moving in unnatural ways without following social protocol, hence forcing people to behave very differently from human-human interaction rules, which can be overcome if robots instead were proactive. In this paper, we build an intention-aware proactive motion planning strategy for mobile robots that coexist with multiple humans. We propose a framework that uses Hidden Markov Model (HMM) theory with a history of observations to: i) predict future states and estimate the likelihood that humans will cross the path of a robot, and ii) concurrently learn, update, and improve the predictive model with new observations at run-time. Stochastic reachability analysis is proposed to identify multiple possibilities of future states and a control scheme that leverages temporal virtual physics inspired by spring-mass systems is proposed to enable safe proactive motion planning. The proposed approach is validated with simulations and experiments involving an unmanned ground vehicle (UGV) performing go-to-goal operations in the presence of multiple humans, demonstrating improved performance and effectiveness of online learning when compared to reactive obstacle avoidance approaches.

This paper presents a novel design concept of a miniaturized Magneto-Rheological (MR) clutch. The design uses a set of spur gears as a means to control the torque. MR clutches with various configurations such as disk-, drum-, and armature-based have in the past been reported in the literature. However, to the best of our knowledge, the design of a clutch with spur gears to use MR fluid in squeeze mode is a novel concept that has never been reported previously.After a brief description of the MR clutch principles, the details of the mechanical design of the spur gear MR clutch are discussed. The distribution of the magnetic flux inside the MR clutch is studied using finite element analysis in COMSOL Multiphysics software. Preliminary experimental results using a prototype MR clutch that validates the new concept and the results therein will be presented next. To clearly show the performance of the proposed design, we compared the torque capacity of our MR clutch obtained experimentally with that of a simulated disk-type MR clutch of a similar size.

This paper presents a trajectory planning algorithm for person following that is more comprehensive than existing algorithms. This algorithm is tailored for a front-wheel-steered vehicle, is designed to follow a person while avoiding collisions with both static and moving obstacles, simultaneously optimizing speed and steering, and minimizing control effort. This algorithm uses nonlinear model predictive control, where the underling trajectory optimization problem is approximated using a simultaneous method. Results collected in an unknown environment show that the proposed planning algorithm works well with a perception algorithm to follow a person in uneven grass near obstacles and over ditches and curbs, and on asphalt over train-tracks and near buildings and cars. Overall, the results indicate that the proposed algorithm can safely follow a person in unknown, dynamic environments.

Compliance is important for humanoid robots, especially a position-controlled one, to perform tasks in complicated environments where unexpected or sudden contacts will result in large impacts which may cause instability or destroy the hardware of robots. This paper presents a compliance control method based on viscoelastic model for humanoid robots to survive on these conditions. The viscoelastic model is used to obtain the relationship between the differential of contact force/torque and linear/angular position. Thus a state equation of this model can be established and a state feedback controller adjusting the position to adapt to the contact force/torque can be designed to realize the compliant movement. The proposed compliance control method based on viscoelastic model has been employed in ankle compliance for stable walking on indefinite uneven terrain and arm compliance for falling protection on BHR-6P, a position-controlled humanoid robot, which validates its effectiveness.

Physical softness has been proposed to absorb impacts when establishing contact with a robot or its workpiece, to relax control requirements and improve performance in assembly and insertion tasks. Previous work has focused on special end effector solutions for isolated tasks, such as the peg-in-hole task. However, as many robot tasks require the precision of rigid robots, and their performance would degrade when simply adding compliance, it has been difficult to take advantage of physical softness in real applications. A wrist that could switch between soft and rigid modes could solve this problem, but actuators with sufficient strength for this state transition would increase the size and weight of the module and decrease the payload of the robot. To solve this problem, we propose a novel design of a soft module consisting of a cable-driven mechanism, which allows the robot end effector to change between soft and rigid mode while being very compact and light. The module effectively combines the advantages of soft and rigid robots, and can be retrofitted to existing robots and grippers while preserving the characteristics of the robotic system. We evaluate the effectiveness of our proposed design through experiments modeling assembly tasks, and investigate design parameters quantitatively.

While abstract knowledge like cause-and-effect relations enables robots to problem-solve in new environments, acquiring such knowledge remains out of reach for many traditional machine learning techniques. In this work, we introduce a method for a robot to learn an explicit model of cause-and-effect by constructing a structural causal model through a mix of observation and self-supervised experimentation, allowing a robot to reason from causes to effects and from effects to causes. We demonstrate our method on tool affordance learning tasks, where a humanoid robot must leverage its prior learning to utilize novel tools effectively. Our results suggest that after minimal training examples, our system can preferentially choose new tools based on the context, and can use these tools for goal-directed object manipulation.

For high-level human-robot interaction tasks, 3D scene understanding is important and non-trivial for autonomous robots. However, parsing and utilizing effective environment information of the 3D scene is not trivial due to the complexity of the 3D environment and the limited ability for reasoning about our visual world. Although there have been great efforts on semantic detection and scene analysis, the existing solutions for parsing and representation of the 3D scene still fail to preserve accurate semantic information and equip sufficient applicability. This study proposes a bottomup construction framework for structured 3D scene graph generation, which efficiently describes the objects, relations and attributes of the 3D indoor environment with structured representation. In the proposed method, we adopt visual perception to capture the semantic information and inference from scene priors to calculate the optimal parse graph. Afterwards, an improved probabilistic grammar model is used to represent the scene priors. Experiment results demonstrate that the proposed framework significantly outperforms existing methods in terms of accuracy, and a demonstration is provided to verify the applicability in applying to high-level human-robot interaction tasks. The supplementary video can be accessed at the following link: https://youtu.be/vEWNxnSwmKI.

While most modern-day quadruped robots crouch their limbs during the stance phase to stabilize the trunk, mammals exploit the inverted-pendulum motions of their limbs and realize both efficient and stable walking. Although the flexibility of the shoulder region of mammals is expected to contribute to reconciling the discrepancy between the forelimbs and hindlimbs for natural walking, the complex body structure makes it difficult to understand the functionality of animal morphology. In this study, we developed a simple robot model that mimics the flexibility of shoulder region in the sagittal plane, and we conducted a two-dimensional simulation. The results suggest that the flexibility of the shoulder contributes to absorbing the different motions between the forelimbs and hindlimbs.

We present a modified TacTip biomimetic optical tactile sensor design which demonstrates the ability to induce and detect incipient slip, as confirmed by recording the movement of markers on the sensor's external surface. Incipient slip is defined as slippage of part, but not all, of the contact surface between the sensor and object. The addition of ridges - which mimic the friction ridges in the human fingertip - in a concentric ring pattern allowed for localised shear deformation to occur on the sensor surface for a significant duration prior to the onset of gross slip. By detecting incipient slip we were able to predict when several differently shaped objects were at risk of falling and prevent them from doing so. Detecting incipient slip is useful because a corrective action can be taken before slippage occurs across the entire contact area thus minimising the risk of objects been dropped.

This paper presents a novel Bayesian-based controller for snake robots in cluttered environment. It extends the conventional shape-based compliant control into statistical field providing an explicit mathematical formulation with Bayesian network. A sequential density propagation rule is derived by introducing several probability densities in a unified framework. Specifically, two input influence densities are proposed to model the cumulative effect of various external forces that the snake robot undergoes. Moreover, the measurement likelihood model is exploited to give a more robust closed-loop feedback. Overall, the proposed approach provides an innovative way to handle challenging tasks of snake robot control in complicated environment. Experimental results have been demonstrated for both simulation and real-world data.

The main contribution of this paper is a probabilistic estimator that assists a mobile robot to locate a gas source in an indoor environment. The scenario is that a robot equipped with a gas sensor enters a building after the gas is released due to a leak or explosion. The problem is discretized by dividing the environment into a set of regions and time into a set of time intervals. Likelihood functions describing the probability of obtaining a certain gas concentration measurement at a given location at a given time interval are assembled using data generated with GADEN, a three-dimensional gas dispersion simulator([1]). Given a measurement of the gas concentration is available, Bayes's rule is used to compute the joint probability density describing the location of the gas source and the time at which it started spreading. To illustrate the estimation process, a relatively simple motion planner that directs the robot towards the most likely gas source location using a cost function based on the marginal probability of the gas source location is used. The motion plan is periodically revised to reflect the latest posterior probability density. Simulation experiments in a large air-conditioned building with turbulence and wind are presented to demonstrate the effectiveness of the proposed technique.

We propose a surface-to-surface (S2S) point registration algorithm by exploiting the Gaussian Process Implicit Surfaces for partially overlapping 3D surfaces to estimate the 6D pose transformation. Unlike traditional approaches, that separate the corresponding search and update steps in the inner loop, we formulate the point registration as a nonlinear non-constraints optimization problem which does not explicitly use any corresponding points between two point sets. According to the implicit function theorem, we form one point set as a Gaussian Process Implicit Surfaces utilizing the signed distance function, which implicitly creates three manifolds. Points on the same manifold share the same function value, indicated as {1, 0, -1}. The problem is thus converted into finding a rigid transformation that minimizes the inherent function value. This can be solved by using a Gauss-Newton (GN) or Levenberg-Marquardt (LM) solver. In the case of a partially overlapping 3D surface, the Fast Point Feature Histogram (FPFH) algorithm is applied to both point sets and a Principal Component Analysis (PCA) is performed on the result. Based on this, the initial transformation can then be computed. We conduct experiments on multiple point sets to evaluate the effectiveness of our proposed approach against existing state-of-the-art methods.