Robotic Perception System Research Articles

Distance-Aware Dynamically Weighted Roadmaps for Motion Planning in Unknown Environments

The paper presents and evaluates a distance-aware dynamic roadmap (DA-DRM) algorithm as an extension of the dynamic roadmap (DRM) approach. In contrast to previous work, the algorithm is capable of planning collision-free trajectories while considering the distance to obstacles, even in unknown environments, which are perceived by the robot's depth camera system. The algorithm makes use of a voxel distance grid, which is updated based on perceptual information acquired from the robot's perception system. The distance information is considered as a cost factor during the roadmap search and it is considered in a postprocessing step that is used for trajectory smoothing. We evaluate the DA-DRM algorithm in simulation and in a real-world experiments with the humanoid robot ARMAR-III. In addition, we compare our algorithm against the DRM and the RRT-Connect algorithm. The results demonstrate the performance of our algorithm in terms of keeping a safety distance to obstacles, trajectory smoothness as well as the ability to generate solutions in narrow free space.

Design of biomimetic human-skin-like tactile flexible sensor

PurposeThe hands of intelligent robots perceive external stimuli and respond effectively according to tactile or pressure sensors. However, the traditional tactile and pressure sensors cannot perform human-skin-like intelligent properties of high sensitivity, large measurement range, multi-function and flexibility simultaneously. The purpose of this paper is to present a flexible tactile-pressure sensor based on hyper-elastics polydimethylsiloxane and plate capacitance.Design/methodology/approachWith regard to this problem, this paper presents a flexible tactile-pressure sensor based on hyper-elastics PDMS and plate capacitance. The sensor has a size of 10 mm × 10 mm × 1.3 mm and is composed of four upper electrodes, one middle driving electrode and one lower electrode. The authors first analyzed the structure and the tactile-pressure sensing principle of human skin to obtain the design parameters of the sensor. Then they presented the working principle, material selection and mechanical structure design and fabrication process of the sensor. The authors also fabricated several sample devices of the sensor and carried out experiments to establish the relationship between the sensor output and the pressure.FindingsThe results show that the tactile part of the sensor can measure a range of 0.05-1N/mm2 micro pressure with a sensitivity of 2.93 per cent/N and a linearity of 0.03 per cent. The pressure part of the sensor can measure a range of 1-30N/mm2 pressure with a sensitivity of 0.08 per cent/N and a linearity of 0.07 per cent.Originality/valueThis paper analyzes the tactile and pressure sensing principles of human skin and develop an intelligent sensitive human-skin-like tactile-pressure sensor for intelligent robot perception systems. The sensor can achieve to imitate the tactile and pressure function simultaneously with a measurement resolution of 0.01 N and a spatial resolution of 2 mm.

Software Product Line Engineering for Robotic Perception Systems

Control systems for autonomous robots are concurrent, distributed, embedded, real-time and data intensive software systems. A real-world robot control system is composed of tens of software components. For each component providing robotic functionality, tens of different implementations may be available. The difficult challenge in robotic system engineering consists in selecting a coherent set of components, which provide the functionality required by the application requirements, taking into account their mutual dependencies. This challenge is exacerbated by the fact that robotics system integrators and application developers are usually not specifically trained in software engineering. In various application domains, software product line (SPL) development has proven to be the most effective approach to face this kind of challenges. In a previous paper [D. Brugali and N. Hochgeschwender, Managing the functional variability of robotic perception systems, in First IEEE Int. Conf. Robotic Computing, 2017, pp. 277–283.] we have presented a model-based approach to the development of SPL for robotic perception systems, which integrates two modeling technologies developed by the authors: The HyperFlex toolkit [L. Gherardi and D. Brugali, Modeling and reusing robotic software architectures: The HyperFlex toolchain, in IEEE Int. Conf. Robotics and Automation, 2014, pp. 6414–6420.] and the Robot Perception Specification Language (RPSL) [N. Hochgeschwender, S. Schneider, H. Voos and G. K. Kraetzschmar, Declarative specification of robot perception architectures, in 4th Int. Conf. Simulation, Modeling, and Programming for Autonomous Robots, 2014, pp. 291–302.]. This paper extends our previous work by illustrating the entire development process of an SPL for robot perception systems with a real case study.

Revisiting active perception

Despite the recent successes in robotics, artificial intelligence and computer vision, a complete artificial agent necessarily must include active perception. A multitude of ideas and methods for how to accomplish this have already appeared in the past, their broader utility perhaps impeded by insufficient computational power or costly hardware. The history of these ideas, perhaps selective due to our perspectives, is presented with the goal of organizing the past literature and highlighting the seminal contributions. We argue that those contributions are as relevant today as they were decades ago and, with the state of modern computational tools, are poised to find new life in the robotic perception systems of the next decade.

Effective Vision‐based Classification for Separating Sugar Beets and Weeds for Precision Farming

The use of robots in precision farming has the potential to reduce the reliance on herbicides and pesticides through selectively spraying individual plants or through manual weed removal. A prerequisite for that is the ability of the robot to separate and identify the value crops and the weeds in the field. Based on the output of the robot's perception system, it can trigger the actuators for spraying or removal. In this paper, we address the problem of detecting sugar beet plants as well as weeds using a camera installed on a mobile field robot. We propose a system that performs vegetation detection, local as well as object‐based feature extraction, random forest classification, and smoothing through a Markov random field to obtain an accurate estimate of crops and weeds. We implemented and thoroughly evaluated our system using a real farm robot in different sugar beet fields, and we illustrate that our approach allows for accurately identifying weeds in a field.

Fitting primitive shapes in point clouds: a practical approach to improve autonomous underwater grasp specification of unknown objects

This article presents research on the subject of autonomous underwater robot manipulation. Ongoing research in underwater robotics intends to increase the autonomy of intervention operations that require physical interaction in order to achieve social benefits in fields such as archaeology or biology that cannot afford the expenses of costly underwater operations using remote operated vehicles. Autonomous grasping is still a very challenging skill, especially in underwater environments, with highly unstructured scenarios, limited availability of sensors and adverse conditions that affect the robot perception and control systems. To tackle these issues, we propose the use of vision and segmentation techniques that aim to improve the specification of grasping operations on underwater primitive shaped objects. Several sources of stereo information are used to gather 3D information in order to obtain a model of the object. Using a RANSAC segmentation algorithm, the model parameters are estimated and a set of feasible grasps are computed. This approach is validated in both simulated and real underwater scenarios.

Towards an Electric-Sense-Based Bioinspired Embodied Robotic Perception System: The Modelling Aspect

In the context of the new paradigm of embodied intelligence invoked by both the roboticians and the cognitive scientists, a sensor bio-inspired from the electric fish was built. A certain geometry was pointed out for an accurate analytical prediction of the electrical measurements on the body in the presence of exterior objects. Such perception model can establish potentially a direct relation between the location of the object and the body positions measurements as well as the shape of the object and the body direction. In addition to a potential novel tomography technique and a new potential electrolocation system it offers new insights in understanding the role of the body in perceiving the world.

MOTION PLANNING USING PREDICTED PERCEPTIVE CAPABILITY

We present an approach to motion planning for humanoid robots that aims to ensure reliable execution by augmenting the planning process to reason about the robot's ability to successfully perceive its environment during operation. By efficiently simulating the robot's perception system during search, our planner utilizes a perceptive capability metric that quantifies the 'sensability' of the environment in each state given the task to be accomplished. We have applied our method to the problem of planning robust autonomous grasping motions and walking sequences as performed by an HRP-2 humanoid. A fast GPU-accelerated 3D tracker is used for perception, with a grasp planner and footstep planner incorporating reasoning about the robot's perceptive capability. Experimental results show that considering information about the predicted perceptive capability ensures that sensing remains operational throughout the grasping or walking sequence and yields higher task success rates than perception-unaware planning.

Robot Perception Systems: Acoustics Vision Set-Up Control

Our work deals with robot perception systems. We address the perception control problem through the study of multisensor cooperation driven by a stochastic controller. One originality of our solution comes from the probabilistic control based on bayesian estimation to positioning the observation windows towards regions of interest. Another one is the asynchronous way to integrate various informations from the different sensors, allowing a best efficiency in the case of disparate costs of numerical processing associated to the sensors. After formulating our perception control model, we describe an acoustics vision set-up and our first experimentations in an indoor environment. Finally, we conclude the proposed paper by presenting the perspectives intended for this multisensor set-up