IROS 2018 Papers

In this paper, we investigate the use of ultra low-power, mixed signal analog/digital neuromorphic hardware for implementation of biologically inspired neuronal path integration and map formation for a mobile robot. We perform spiking network simulations of the developed architecture, interfaced to a simulated robotic vehicle. We then port the neuronal map formation architecture on two connected neuromorphic devices, one of which features on-board plasticity, and demonstrate the feasibility of a neuromorphic realization of simultaneous localization and mapping (SLAM).

Robots deployed for long periods of time need to be able to explore and learn from their environment. One approach to this problem has been reinforcement learning (RL), in which robots receive rewards from the environment that allow them to choose optimal actions. To speed learning when human supervision is available, interactive reinforcement learning solicits feedback from a human teacher. However, this approach typically assumes that learning takes place under continuous supervision, which is unlikely to hold in long-term scenarios. We propose an extension to a method of interactive reinforcement learning, policy shaping, that takes into account human attention. Our approach enables better performance while unattended by favoring information-gathering actions when attended and actions that have received positive feedback when unattended. We test our approach in both simulation and on a robot, finding that our method learns faster than policy shaping and performs more safely than policy shaping while no one is paying attention to the robot.

Skin impedance should be minimized to obtain reliable and precise surface electromyography (sEMG) signals. High skin impedance decreases sensitivity to muscular activation and makes sEMG signals vulnerable to external noise. Microneedle-based electrodes have been proposed to achieve low skin impedance and high spatial resolution. However, unstable skin contact can occur during dynamic motion due to the electrodes small contact area, and the signal is easily influenced by motion artifacts. In this study, a pneumatic microneedle-based high-density sEMG sleeve is proposed that guarantees stable sEMG signal measurement by pressing the electrodes to the skin. Pneumatic air control, sEMG signal processing, and wireless signal transmission are processed in a single processor. The proposed interface automatically controls the air volume according to sEMG signal quality and is comfortable for users. The usability of the proposed interface was compared to conventional Velcro armbands by examining acquired sEMG signals. The results indicated that the proposed pneumatic sleeve guarantees reliable sEMG signal measurement during dynamic motion. Additionally, optimal pneumatic pressure and needle length were investigated.

Indoor localization is critical to many location-aware applications, however, a low-cost solution with guaranteed accuracies has not yet come. Visible Light Communication (VLC-) based localization techniques are very promising to fill this gap. In this paper, we propose Plugo, a novel VLC system with random multiple access towards low-cost indoor localization. Compared to conventional RF-based approaches that rely on dedicated wireless access points as location beacons, the proposed system has the potential to deliver better accuracies with reduced cost. Specifically, we build a handful of compact VLC-compatible LED bulbs out of low-cost offthe-shelf components (around $10 total cost for each assembly) and recover VLC signals using a cheap photodiode receiver. The basic framed slotted Additive Links On-line Hawaii Area (ALOHA) is exploited to achieve random multiple access over the shared optical medium. We show its effectiveness in beacon broadcasting by experiments, and further, demonstrate a preliminary localization result with sound accuracy by using fingerprinting-based methods in a customized testbed.

In order to fully function in human environments, robot perception needs to account for the uncertainty caused by translucent materials. Translucency poses several open challenges in the form of transparent objects (e.g., drinking glasses), refractive media (e.g., water), and diffuse partial occlusions (e.g., objects behind stained glass panels). This paper presents Plenoptic Monte Carlo Localization (PMCL)as a method for localizing object poses in the presence of translucency using plenoptic (light-field)observations. We propose a new depth descriptor, the Depth Likelihood Volume (DLV), and its use within a Monte Carlo object localization algorithm. We present results of localizing and manipulating objects with translucent materials and objects occluded by layers of translucency. Our PMCL implementation uses observations from a Lytro first generation light field camera to allow a Michigan Progress Fetch robot to perform grasping.

Provides an abstract for each of the plenary presentations and may include a brief professional biography of each presenter. The complete presentations were not made available for publication as part of the conference proceedings.

We present an approach to plan trajectories for a fleet of fixed-wing UAVs to observe a wildfire evolving over time. Realistic models of the terrain, of the fire propagation process, and of the UAVs are exploited, together with a model of the wind. The approach tailors a generic Variable Neighborhood Search method to these models and associated constraints. Simulation results show ability to plan observation trajectories for a small fleet of UAVs, and to update the plans when new information on the fire are incorporated in the fire model.

The ability to move around the environment is one of the most important capabilities of a mobile robot. Although navigation is considered an already achieved capacity, there is still much work to be done to integrate navigation with high level reasoning and acting. Navigate in indoor environments also involve complex actions, such as opening doors, use elevators, and many others. We propose a topological navigation system based on Artificial Intelligence (AI)Planning. Starting from a symbolic representation of the environment, navigation tasks are divided into phases, in which different actions are required. This approach has demonstrated to be very effective to plan the operations of a robot at indoor environments. The final result is method compact, efficient and scalable. Our system has been successfully tested at European Robotics League in the humanoid robot Pepper.

This paper addresses the problem of obtaining human-like motions on hand-arm robotic systems performing grasping actions. The focus is set on the coordinated movements of the robotic arm and the anthropomorphic mechanical hand, with which the arm is equipped. For this, human movements performing different grasps are captured and mapped to the robot in order to compute the human hand synergies. These synergies are used to both obtain human-like movements and to reduce the complexity of the planning phase by reducing the dimension of the search space. In addition, the paper proposes a sampling-based planner, which guides the motion planning following the synergies and considering different types of grasps. The introduced approach is tested in an application example and thoroughly compared with a state-of-the-art planning algorithm, obtaining better results.

The use of robots, as a social stimulus, provides several advantages over using another animal. In particular, for rat-robot studies, robots can produce social behaviour that is reproducible across trials. In the current work, we outline a framework for rat-robot interaction studies, that consists of a novel rat-sized robot (PiRat), models of robotic behavior, and a position tracking system for both robot and rat. We present the design of the framework, including constraints on autonomy, latency, and control. We pilot tested our framework by individually running the robot rat with eight different rats, first through a habituation stage, and then with PiRat performing two different types of behaviour - avoiding and frequently approaching. We evaluate the performance of the framework on latency and autonomy, and on the ability to influence the behaviour of individual rats. We find that the framework performs well on its constraints, engages some of the rats (according to the number of meetings), and features a control scheme that produces reproducible behaviour in rats. These features represent a first demonstration of a closed-loop rat-robot framework.

A compelling robotics course begins with a compelling robot. We introduce a new low-cost aerial educational platform, the PiDrone, along with an associated college-level introductory robotics course. In a series of projects, students incrementally build, program, and test their own drones to create an autonomous aircraft capable of using a downward facing RGB camera and infrared distance sensor to visually localize and maintain position. The PiDrone runs Python and the Robotics Operating System (ROS) framework on an onboard Raspberry Pi, providing an accessible and inexpensive platform for introducing students to robotics. Students can use any web and SSH capable computer as a base station and programming platform. The projects and supplementary homeworks introduce PID control, state estimation, and high-level planning, giving students the opportunity to exercise their new skills in an exciting long-term project.

This paper presents a visual odometry approach using a Pixel Processor Array (PPA) camera, specifically, the SCAMP-5 vision chip. In this device, each pixel is capable of storing data and performing computation, enabling a variety of computer vision tasks to be carried out directly upon the sensor itself. In this work the PPA performs HDR edge detection, perspective correction and image alignment based odometry, allowing the position and heading of a MAV to be tracked at several hundred frames per second. We evaluate our PPA based approach by direct comparison with a motion capture system for a variety of trajectories. These include rapid accelerations that would incur significant motion blur at low frame rates, and lighting conditions that would typically lead to under or over exposure of image detail. Such challenging conditions would often lead to unusable images when relying on traditional image sensors.

In this paper we present a data driven approach for safe and smooth autonomous navigation of a personal mobility vehicle (PMV) when facing moving obstacles such as people and bicycles in public pedestrian paths. In a period of three months, data from five different persons driving the robotic PMV in an outdoor environment while facing pedestrians were collected. 2465 clean tracks around the vehicle together with PMVs trajectories were collected. We performed an analysis of the parameters involved for human-driven smooth navigation. Relevant parameters regarding PMV-Human interaction included distance to moving objects, passing side and velocities. Moreover, data suggests the existence of a social navigational distance for the PWv. For autonomous navigation we implemented a Frenet planner to achieve safe and smooth navigation for the passenger and pedestrians around. Experimental results in real pedestrian paths show that the PMV is capable of smoothly following its path while facing pedestrians and bicycles.

There are many applications such as surveillance and mapping that require persistent monitoring of terrains. In this work, we consider a heterogeneous team of aerial and ground robots that are tasked with monitoring a terrain along a given path. Both types of robots are equipped with cameras that can monitor the terrain within their fields-of-view. We also consider the ability of the aerial robots to land occasionally on the terrain to recharge. The objective is to find a path for all the robots to reduce the time required. Determining optimal routes for the robots is a challenging problem because of constrained visibility due to the terrain and fuel limitations of the robots. We devise an MILP formulation for the problem using a 1.5 dimensional representation model. A branch-and-cut framework is used to implement the MILP and involves the design of a separation algorithm to compute valid inequalities. We report results from extensive simulations and proof-of-concept field experiments to show the efficacy of our approach.

We present a fast and very effective method for object classification that is particularly suited for robotic applications such as grasping and semantic mapping. Our approach is based on a Random Forest classifier that can be trained incrementally. This has the major benefit that semantic information from new data samples can be incorporated without retraining the entire model. Even if new samples from a previously unseen class are presented, our method is able to perform efficient updates and learn a sustainable representation for this new class. Further features of our method include a very fast and memory-efficient implementation, as well as the ability to interrupt the learning process at any time without a significant performance degradation. Experiments on benchmark data for robotic applications show the clear benefits of our incremental approach and its competitiveness with standard offline methods in terms of classification accuracy.

The direct kinematics problem of the Stewart-Gough platform can be solved by measuring the manipulator platform's orientation and two of the linear actuators' orientations instead of the six linear actuators' lengths. In this paper, the effect of measurement errors on the calculated manipulator platform's pose is investigated using the Cramer-Ran lower bound and extensive experiments on a state-of-the-art Stewart-Gough platform. Furthermore, different algorithms and filters for one-time as well as continuous pose determinations are investigated. Finally, possible sensor fusion concepts for the one-time pose determination are presented to increase the robustness against noise and measurement errors.

Most existing motion planning algorithms assume that a map (of some quality) is fully determined prior to generating a motion plan. In many emerging applications of robotics, e.g., fast-moving agile aerial robots with constrained embedded computational platforms and visual sensors, dense maps of the world are not immediately available, and they are computationally expensive to construct. We propose a new algorithm for generating plan graphs which couples the perception and motion planning processes for computational efficiency. In a nutshell, the proposed algorithm iteratively switches between the planning sub-problem and the mapping sub-problem, each updating based on the other until a valid trajectory is found. The resulting trajectory retains a provable property of providing an optimal trajectory with respect to the full (unmapped) environment, while utilizing only a fraction of the sensing data in computational experiments.

In order to accurately reach a target position while executing a task which imposes occlusion or constraints of the posture, a humanoid robot requires an adaptive locomotion system, which can comprehensively integrate localization, environmental mapping, global locomotion planning and local error correction. In this paper, we propose a method of constructing a perception based locomotion system for a humanoid robot. The major contribution of this paper is solving a problem of the locomotion error caused by the task constraints, by locally compensating footsteps and assessing the need for global footstep re-planning online based on environmental measurements. The proposed system provides an accurate and dense ground point cloud, called HeightField, using plane estimation and space interpolation, and obstacle point cloud for frequent collision avoidance by accumulating laser scans. This environmental perception enables a humanoid robot to plan footsteps globally even in the situation where the sight of the robot is limited and compensate footsteps while estimating landing state during locomotion online with the localization result. We evaluated the practicality of the proposed system by applying it to our humanoid robot carrying a heavy object in a construction site and confirmed that the proposed system contributed to improved locomotion abilities of a humanoid robot engaging in heavy-duty or dangerous tasks.

Despite growing attention in autonomy, there are still many open problems, including how autonomous vehicles will interact and communicate with other agents, such as human drivers and pedestrians. Unlike most approaches that focus on pedestrian detection and planning for collision avoidance, this paper considers modeling the interaction between human drivers and pedestrians and how it might influence map estimation, as a proxy for detection. We take a mapping inspired approach and incorporate people as sensors into mapping frameworks. By taking advantage of other agents' actions, we demonstrate how we can impute portions of the map that would otherwise be occluded. We evaluate our framework in human driving experiments and on real-world data, using occupancy grids and landmark-based mapping approaches. Our approach significantly improves overall environment awareness and outperforms standard mapping techniques.