Mobile Service Robotics

The following sections are included:

• TITLE
• PREFACE
• CONFERENCE ORGANISERS
• CONFERENCE COMMITTEES AND CHAIRS
• CONFERENCE SPONSORS AND CO-SPONSORS
• TABLE OF CONTENTS

The synchronization of legs in many-legged robots is a combinatorial problem that can be very critical for successful climbing or locomotion in very rough terrain. This paper addresses developing gait generation controllers with desirable synchronization properties by introducing a number of abstractions in the traditional modeling approaches. It describes how to use the max-plus algebra as a modeling tool for discrete-event systems and how it can be applied to legged locomotion.

Note from Publisher: This chapter contains only abstract.

Note from Publisher: This chapter contains only abstract.

Since the 1960's, robotics has focused on manufacturing robots for use in industrial applications to perform operations such as handling, joining, inspection, machining, spot welding, assembling, etc. In order for the robots to perform their required tasks in these classical applications, the “workpiece” has had to be brought to the robots. Since the mid-1980s, interest has grown in developing new robots for applications where tasks need to be performed in-situ locations demanding that problems of localisation and locomotion are also important. In these scenarios, robot mobility is the key capability, and this can be realised in a variety of ways, e.g., via wheels, tracks, legs, etc. In some applications, the mobile robots also need a climbing capability due to their working environments and tasks to be performed (e.g., inspecting the external walls of tall buildings); these requirements present interesting challenges as seen in the area of climbing and walking robots focussed on by the EC funded Network of Excellence CLAWAR (climbing and walking robots) coordinated by Professor Virk during 1996-2005. Researchers have used CLAWAR to discuss climbing AND walking AND running robots but in fact any form of robot mobility is relevant and presents interesting challenges for robotics…

In order to get symmetric and natural gait for prosthesis wearer, a mapping control method is presented based on echo control method. Firstly, state phase division method by using angle sensor is developed. Then according to the characteristic of each state phase, Angle-mapping model(Angle as main parameter of the model) is built by use of curve fitting method. Furthermore, according to human motion measurement data, control simulation based on the mapping model is performed to verify the correctness of the model. Lastly, the adaption ability to velocity is analyzed for the model. And analysis result shows that the presented model can adapt a fixed velocity with a few errors.

Low volume industrial productions are rarely highly automated because of the related costs. Variable production requires flexible automation with close human robot interaction. An exoskeleton may exactly provide these features to enhance industrial production. This article highlights the difficulties related to using exoskeletons in an industrial setting. Moreover, it introduces the Robo-Mate project – an EU funded project – targeted to address the application of an exoskeleton in industry.

In our current research, we have developed a novel standing assistance system for an elderly person. In general, many previous assistive devices do not require a patient to use own physical strength and it causes the decreasing the physical strength of an elderly people. Therefore, a standing assistance system, which uses a remaining physical strength of a patient, is required. For realizing it, the assistance system needs to estimate the load of a patient during standing motion and in many cases, they use a joint traction which can be derived by an human linkages model kinematically for evaluating the load of a patient. However, human body motions are generated by many muscles and a joint traction shows only the body motion. This means it is difficult to evaluate the muscular power of a patient during a standing motion using a joint traction and it is not a suitable index for a load estimation of the patient. Thus, in this paper, we propose the load estimation scheme which consider the muscle functionality during standing motion of the patient. For realizing it, our proposed system estimates the muscular power using the human musculo-skeletal model of a lower limb which considers a biarticular muscle function. Using its estimated muscular power, our system evaluates the patient's load at real time. Our idea considers the characteristics of the muscular power which changes according to the posture of a patient and we verify that estimated results fits opinions of nursing specialists and the muscle activation level derived by EMG data.

This paper presents a hybrid approach involving geometric and temporal evolution of joint motion for the imitation of human motion by a humanoid robot. We exploit the key advantages of modifying temporal evolution of joint motion like joint coordination and Cartesian trajectory tracking and combine it with geometric evolution of joint motion to ensure balance and obey the physical limits of the humanoid robot during motion imitation. In order to achieve this objective, we have framed a multi objective optimization problem with constraints on balance. A feasible set of joint motion was obtained, as close as possible to the human reference motion, to achieve a natural looking motion imitation with optimum time delay and joint angle error. This approach was tested with the NAO humanoid robot with a kick motion in a single-support phase.

This paper presents the design of a gait restoration system which maximises locomotion of lower limbs by means of the user's own muscle contraction, and reduces the number of sensors and actuators to minimise power consumption. The system comprises a computer aided design of a humanoid model, an orthotic exoskeleton device and a wheel walker within a virtual environment. The system is later controlled using fuzzy logic to generate knee flexion by applying functional electrical stimulation (FES) to a hamstrings muscle model. It is coupled with a finite state controller (FSC) that activates brakes aligned with the lower limbs joints in the sagittal plane to generate specific postures needed to produce a continuous walking cycle. Simulations are run as initial validation of the system.

The paper presents a research update on the AAL Call 4 EXO-LEGS project aimed at developing lower-body mobility exoskeletons to assist elderly persons to stay independent in their normal daily living activities for as long as possible. The important movement functionalities and key design issues to be included in the process are identified via specifically developed questionnaires and responses from a pan-European end user group set up as part of the project. The user requirements are used together with the recently published ISO safety requirements for personal care robots to perform targeted technical research in the areas of human gait analysis, modelling and simulation, mechanical engineering, embedded system design, and ergonomic user interfacing.

Two important aspects of today's production are energy and time consumption, a typical production process requires parts to be constantly moved between locations. The aim of this research is to minimize the energy and time consumption during the assembly and transport phases by finding the optimal path for the products to be moved. This paper studies trajectory planning and minimization for mobile robots. The trajectory is generated using function approximations and the robot behavior is studied among different trajectories to find the best algorithms to minimize time and energy consumption. The methods discussed in this paper could be repeated automatically several times for better minimization results. The focus of this paper is mobile robotics usage in inspection, industrial and agricultural environments.

In this paper the settlement scheme of the four-link jumping robot is submitted, the mathematical model of movement of object in which the jump of the device is presented in the form of sequence of phases is developed, dependences between the jump characteristics and the position of the point of fixing of a foot are received.

Modular robots are highly flexible systems that are able to reconfigure to various shapes and to perform diverse locomotions. To visit a desired place in a complex environment, a robot can be equipped with several motion primitives that are switched using a high-level motion planning. The utilization of the high-level motion planning allows to simplify the motion primitives, which only need to provide short motions in a vicinity of the robot. This paper investigates how to automatically optimize these motion primitives of modular robots using an evolutionary approach. The proposed optimization is experimentally verified in simulated scenarios and in experiments with real robots.

A new intelligent wall-climbing welding robot system (IWWRS) is presented in this paper, which is composed of robot body, control system and welding system. The robot body including a mobile platform and a three DOFs manipulator is designed by theory analysis of the mechanical model. IWWRS employs a device combined of arrayed gapped adsorption and magnetic-wheel-adsorption to climb on curved ferromagnetic surfaces. The control system is composed of hardware, software and coordinated control algorithm. A set of experiments are carried out to demonstrate and verify the properties of IWWRS in the on-site welding factory. And the application results show that IWWRS can satisfy the demands of automatic welding of large-scale ferromagnetic structure and work with better quality and efficiency compared to manual welding.

This paper presents a system that allows for semi-autonomous decontamination and radiation measurement of surfaces as found during dismantling of nuclear power plants. The hardware setup comprises a transport system and a manipulator equipped with vacuum suction units to move along walls. Furthermore, the algorithms for acquiring and processing an environment model as well as path planning based on the kinematic constraints of the manipulator are discussed. The software framework also allows for seamless switching between simulation and reality for real-time monitoring.

A new structure of two-wheeled vehicle with an extendable intermediate body (IB) is investigated in this paper. The robot design offers an additional feature to the conventional inverted pendulum on two wheels. The IB of the robot is composed of two co-axial parts connected by a linear actuator and with a payload attached to the end of the upper part. The linear actuator allows the payload to move up and down along the IB of the robot. Considering various positions, speeds and different sizes of a payload, carried by the robot, while maintaining upright balance is the main objective of the current study. To develop a robust vehicle that is able to drive and manoeuvre on irregular terrains is an important aspect of this field. The ability of the vehicle to adapt to different terrains allows a variety of mobility solutions to be developed using the vehicle as a basis. Furthermore, the simulations would give a measure of how robust is the vehicle to work on different environments in the real life. In this paper, various simulation scenarios are presented that incorporates the variation of the surface characteristics in terms of the friction and payload movement behaviour.

This paper is concerned with the balancing of a two-wheeled inverted pendulum robot prototype, designed and developed in the University of Sheffield, on an inclined surface. The controller is realized by two PID based control loops, which are designed for balancing and stabilizing the robot considering the tilt angle and wheels speed as feedback. Tilt angle sensing error is reduced by adopting a sensor fusion technique realized by implementing a Kalman filter, which combines the gyroscope and the accelerometer sensor measurements to estimate accurate angle information. The designed controller of the robot is implemented using an ARM microcontroller for real time operation. The robot is tested for balancing the robot on a 60 cm inclined surface with three different inclination angles. The experimental results show that the robot has the ability to maintain stability and balance on an inclined surface.

Path planning and autonomous navigation algorithms play a vital role in the field of robotics. Amongst these, the potential field algorithm is widely used due to its elegant mathematical model. Although it serves the basic purpose of avoiding obstacles, it is bounded by particular restrictions. The use of a virtual obstacle along with potential field algorithm is a lucrative approach to overcome these limitations. This work aims at optimizing certain parameters involved in the virtual obstacle concept by the use of Non-Dominated Sorting Genetic Algorithm II (NSGA II). It is advisable to maintain a safety margin around the obstacle and to maneuver efficiently without oscillations as it moves close to the obstacle. Furthermore, the size of the robot also affects its motion. This paper takes into account all these factors during the optimization process. The results have proven its feasibility and validity in unknown environments.

This article explores the dynamics and control of a six-wheeled mobile robot designed to walk on rough surfaces and considers the synthesis of its control. This robot has pneumatic actuators for lifting the legs and wheel rotation. Each leg moves in the plane and it is independently controlled by two pneumatic cylinders. The six-wheeled robot dynamics model is proposed and studied in this paper. Also, the algorithm of overcoming big obstacles is constructed, which is based on a software package called “Universal Mechanism” (UM).

Significant progress in the design of autonomous robots was achieved not only due to the technological development, but also due to application of biologically inspired methods in the design and control. Advanced and efficient methods of robots motion generation refer to the biological and neurological backgrounds. This especially concerns the walking machines. The development of those robots often requires understanding of biomechanics and neurology of biological systems which they imitate. The historical background concerning biological motion imitation is outlined. Biological principles of motion generation and selected formal models describing animal motion principles, which successfully support motion generation of walking robots are presented. Recent development trends are summarized.

Recently, many researchers are focusing on the human foot mechanism, which has two longitudinal and one lateral arches, to build a new foot mechanism for legged robots. Such research, however, only focuses on the longitudinal effects on the robots, such as their pitching and kicking motions; the lateral effects are omitted (e.g., rolling motion or lateral posture). This paper presents a new foot mechanism inspired by the human foot mechanism with one longitudinal arch and two longitudinal arches that contains mechanical springs. We built a simple biped robot with only one D.C. motor using Chebyshev linkages for the leg mechanism to which we attached our new feet. Through experiments we confirmed that the one-arch mechanism absorbed the shock at the touch down. It also reduced the peak value of the roll angle of the robot body caused by the falling down motion in the lateral direction.

Rapid reacting motions are important for robots to adjust to changes in the real world, which is constantly changing drastically. Pneumatic artificial muscles (PAMs) have been used as actuators of fast-moving robots. Robots equipped with PAMs can move adaptively and dynamically using passive properties of PAMs. In this study, we considered preparation for fast reacting motions by robots equipped with joints driven by competing PAMs (antagonistic PAM-driven joints) as using active properties of PAMs. Specifically, we investigated a method of adjusting initial pressures in PAM-driven joints increasing the maximum joint velocity generated in a short time. We found the optimal initial pressure in PAM-driven joints considering a mechanical delay of PAMs in simulations. Additionally, we verified the validity of results in the simulations using an actual robotic arm.

Recently, many omnidirectional mobile robots have been developed, which require omnidirectional mobile mechanisms to allow them to move within narrow, complicated passages. However, existing omnidirectional mobile mechanisms cannot achieve stable movement because of their small contact areas. In this paper, we propose an omnidirectional mobile robot that achieves stable movement through a spiral-type traveling-wave-propagation mechanism based on a snail's locomotion mechanism. To demonstrate the robot's performance, we conducted a driving experiment using a prototype.

In this work, a self-adjustable transducer for the measurement of strain in semi-stiff leg segments of a six-legged robot is introduced. Appropriate locations and orientations of appendant strain-gauges are discussed based on exoskeletal mechanoreceptors found within the femora and tibiae of stick insects. The transducer system is tested in a single walking leg of the HECTOR robot. Finally, a general scheme of periodic re-calibration is suggested based on the behaviour of biological strain (force) sensors during vertebrate and insect walking.

The 8 DOF quadruped robotic platform is designed and constructed for ground surveillance purpose. Designed model is simulated in physics environment by considering various physical parameters like weight of the robotic platform, torque of actuators, effect of collision. This paper also explains motion planning of a quadruped robotic platform by means of various walking patterns. Walking patterns are tested on different virtual terrains which helps to achieve reliable algorithm and modify robot as per dynamics. The agility and sustainability of robot is employed by biomimetic locomotion. Lateral leg insect is imitated by two DOF leg configuration, which enables basic motion of lifting and shifting. The basic prototype is realized by using eight servo actuators and it is tested on various surfaces having avoidable obstacles.

Although considerable research has been done on motion planning based on demonstrated examples, the ability to generalize from few examples is still an open problem. In this paper, we examine the ability of reproduction and generalization of human-like movements based on the paradigm of dynamic motion primitives (DMPs). The purpose of the current study is to accurately portray the characteristics and generalization performance of this approach, both using simulated and human motion capture data. To this aim, we consider discrete movements that require reach actions toward a target. The adaptation of learnt movements to new situations is performed in the joint and the task spaces. The generalization performance is evaluated and the feasibility of the approach is discussed for this study of reaching movements.

An understanding of the gross mechanics of running is essential for the design of running robots that use dynamically stable gaits. In earlier papers [1, 2] the author and his colleagues analyzed the complete stride cycle for both transverse and rotary gallops. This resulted in a solution that required that the durations of the two flight phases should be equal, in both cases. Examination of experimental results indicates that this conclusion is quite wrong. Review of the analysis indicates that this result was driven by an assumption that the system behaves as a rigid body for motion about the roll axis. Abandoning that assumption produces a simpler analysis which produces results that are broadly consistent with available experimental data.

Within this paper, several robot organisms assembled from multiple CoSMO modules are presented. CoSMO is a mobile modular self-reconfigurable robot platform, capable of working autonomously on its own or connected to other modules building a robot organism. The paper provides an introduction of the possibilities and the diversity of the CoSMO platform and the resulting capabilities to solve different tasks depending on the circumstances.

For their small size, good mobility and energy saving, miniature mobile robots have possibility application in many domains, but it is hard for traditional locomotion mechanisms to fit miniature robot well because of their complex structures. To solve this problem, a kind of resonant locomotion principle based on impact force is proposed and analyzed through numerical simulation of the vibro-impact system model. And then, a prototype locomotion mechanism and an energetically autonomous robot actuated by piezoelectric bimorphs are developed and tested. Experiment results show that the mechanism can achieve locomotion velocity as high as 280mm/s with the 10V excitation voltage, and can also generate motion about 16mm/s with excitation voltage as low as only 1V.

This paper describes the development of Moorinspect, a novel robot that can climb on platform mooring chains both underwater and in air to non-destructively test (NDT) each link with long range ultrasound guided waves. The prototype robot is designed to be able to climb up/down a mooring chain for up to twenty metres below the surface, climb up through the splash zone to the first link which is located in air and connected to the turret of a Floating Production Storage and Offloading (FPSO) facility. This first link suffers the most intense stresses and fatigue failure and thus it is important to test its integrity. A robot that can climb through the splash zone (considered to be the most dangerous zone for human divers) will be the first of its kind. The robot is able to cope with link dimension variations due to corrosion and biological fouling and link curvatures caused by bent links, chain curvature due to gravity, and links twisted at angles of up to eight degrees around a nominal angle of ninety degrees. The robot places an NDT collar consisting of ultrasound probes around the full circumference of each link to be tested. The design has been analysed extensively using Von Mises stress analysis to ensure that the robot is strong and robust enough to carry a sensor payload of more than 70 kg and a robot weighing 500 kg in air and is able to operate in the splash zone. A first prototype chain climbing robot has been tested via trials on a four link mooring chain suspended in air and while immersed in a diving tank.

Friction of sealings is a general problem for sliding wall-climbing robots using negative pressure adhesion for attraction. Tight sealings are very leak-proof but produce higher friction which has to be overcome by the locomotion system, often in terms of tracks or wheels. On the other hand loose sealings could lead to a fail of adhesion and therefore to a drop-off. This paper presents a method to optimize friction characteristics online depending on the current situation without influencing the attraction forces. The approach makes use of an inflatable and controlled rubber sealing and adjusts the air pressure via an overlaying friction controller. Experiments on the new climbing robot CREA prove the functionality and the benefit of the developed method.

Thanks to recent advances in adaptation algorithms, it is now possible to give robots the ability to discover by trial-and-error the best way to behave in unexpected situations, instead of relying on contingency plans. In this paper, we describe the kinematic and mechatronic design of the Creadapt robot, a new mobile robot designed to take full advantage of these recent adaptation algorithms. This robot is a versatile wheel-legged hexapod designed for both legged and wheeled locomotion. It is reversible to be able to continue its mission if it flips over and it uses 6 legs to be able to move efficiently if one or several legs break. This robot also embeds two RGB-D cameras to estimate its velocity onboard, thanks to a RGB-D visual odometry algorithm. Overall, the Creadapt robot is one of the first mobile robot designed with adaptation algorithms in mind.

This paper presents a novel design for a modular snake-like robot with magnetic coupling between the modules. Each module combines 3D-printed mechanical parts with widely available standard electronic components, resulting in a highly customizable, low-cost robot platform for research and education. Since simulation and 3D-printing rely on the same model-files, the design also integrates smoothly into simulation environments like OpenRAVE. The robot can be assembled and re-assembled on the fly. Automatic topology detection is realized with custom connection-interfaces and dynamic initialization of intermodule communication.

The technical capabilities of the transfer of mechanical energy between the different nodes are often limited. There are different mechanisms in which the working body is movable but the drive for constructional reasons must be stationary (e.g. hand devices, robotic joints, etc.). Alternative way to transfer of mechanical energy from the remote drive to the working body is the using of flexible shafts, cables, pneumatic and hydraulic machines, etc. This article describes a pilot version of an electrohydraulic linear drive with a flexible transmission.

A mobile manipulator platform for testing of an intelligent prodder is presented. This instrument permits to give information about the material of a touched unknown object. The recognition is based on the analysis of the contact between the sensing unit and the material. The material is excited with a piezoelectric actuator and its response measured with another piezoelectric sensor. A probe prototype has been designed and built and experimental trials have been performed mounting the system on the mobile manipulator.

Although prodders are still one of the most important tools for humanitarian demining, prodding of antipersonnel mines is considered a risky procedure due to the significant number of accident reported during its practice. With the aim of reducing demining accidents, this paper proposes an intelligent feedback prodder that provides to the deminers information about the amount of force exerted and alerts him when the prodder's angle is approaching or exceeding a certain limit. Results of preliminary tests show the feasibility and reliability of the proposed design and highlight the potential benefits of the tool.

This paper introduces a method to simultaneously optimize design and control parameters for legged robots to improve the performance of locomotion based tasks. The morphology of a quadrupedal robot was optimized for a trotting and bounding gait to achieve a certain speed while tuning the control parameters of a robust locomotion controller at the same time. The results of the optimization show that a change of the structure of the robot can help increase its admissable top speed while using the same actuation units.

Passive Dynamic Walking, which shows natural and efficient gait on shallow slope, makes people believe that biped walking can be considered as the alternate free-rotation of double pendulums. To preserve this property of freerotation and adjust the walking gait for proper step length and period, we consider a novel method to realize biped walking on level ground using coupled elastic actuation. First we introduce the walking model and the control law for the actuated bars, and show the stable walking gait by numerical simulation. Then we analyze the influence of control and physical parameters on walking performance. In addition, an alternative implementation method by coupled elastic actuation on the hip is investigated.

The plantar area of the human foot is larger than that of the quadruped foot, and it contains a large number of sensory organs. Thus, such a foot structure plays a crucial role in extracting “rich” sensory information for the generation of adaptive walking in humans. Here, we propose novel central pattern generator (CPG)-based control of a bipedal walking robot by exploiting plantar sensation. To effectively exploit plantar sensory information, we redesign the local sensory feedback to the CPG model that we previously proposed for quadruped robots. The simulation results indicate that the biped model exhibits a remarkably robust walking ability by exploiting the plantar sensation according to the current walking motion.

This paper proposes an analysis of the effect of vertical motion of the CoM during humanoid walking. The linear inverted pendulum (LIP) model is classically used to deal with humanoid balance during walking. The effects on energy consumption of the CoM height remaining constant for humanoid robots, or varying like human beings, are badly known and are studied here. Two approaches are introduced for the comparison: the LIP which offers the great advantage of analytical solving (i.e. fast and easy calculations), and a numerical solving of the IP dynamics, which allows varying the height of the center of mass during walking. The results are compared in a dynamics simulation of a 2D walking biped.

In this paper we show that exchanging curved feet and rigid ankles by flat feet and compliant ankles improves the range of gait parameters for a bipedal dynamic walker. The new lower legs were designed such that they fit to the old set-up, allowing for a direct and quantitative comparison. The dynamic walking robot RunBot, controlled by an reflexive neural network, uses only few sensors for generating its stable gait. The results show that flat feet and compliant ankles extend RunBot's parameter range especially to more leaning back postures. They also allow the robot to stably walk over obstacles with low height.

This paper presents a limit cycle walking model with segmented feet. We add adaptable ankle and joint stiffness to a limit cycle walker based on previous biological studies to further understand the effects of segmented feet on walking performance. The equilibrium position and stiffness of the toe joints vary during one step depending on the current walking phase. Both the walking phase sequence and the toe joint behavior show a great resemblance to human normal walking. Experimental results show the effects of toe parameters on walking speed, step length, energetic efficiency and the net works of ankle and toe joints, which could provide us a better understanding of segmented-foot limit cycle walkers and help in building a physical robot with toe joints.

Compared to other robotic fields leg motion trajectories of multi-legged walking robots are often quite simple. Although, many walking robots have three or more joints per leg, they do not use their kinematic potential. In this work we propose an adaptive leg trajectory model as basis for a target-oriented trajectory optimization. Together with an interactive graphical editor the leg trajectory model allows robot developers to improve the energy efficiency, stability and velocity of walking robots. All presented components were successfully evaluated in simulation and on the six-legged walking robot LAURON IVc.

Swing-leg retraction, the backward rotation of the swing leg just prior to ground contact, is observed in human locomotion. While several advantages of swing-leg retraction, like gait stability and perturbation rejection, are shown by conceptual models, there is currently very little experimental data on swing-leg retraction in human motion. In this paper, kinematic data for twenty-eight subjects walking and running at different speeds are analyzed. Swing-leg retraction was shown to exist in walking and running at every non-zero speed. Additionally, swing-leg retraction speed and acceleration linearly increase with gait speed. At comparable gait speeds, swing-leg retraction speed is higher for running than for walking.

An omni-directional walking control method for a six-legged robot is established by considering the dynamics of the actuator. We constructed a mathematical model in which the inputs were the driving torque of the rotating links in the support legs and the outputs were the x, y-positions of the body and the yawing angle. An LQI control system was designed. By using a 3D model of the six-legged robot, the effectiveness of the proposed control method was examined in the case that the reference orbit was a semicircle of 0.5 m in radius. As a result, the validity of the proposed control method was confirmed.

Development of mobile robots capable of working in various scenarios such as rescue and exploration has recently gained interest. Because such scenarios often involve irregular terrain, we focus our attention on achieving stable operation of a hexapod robot on irregular terrain. To do so, we developed a leg using a parallel linkage mechanism and artificial muscles made of pneumatic rubber and constructed a model of this leg. The model reproduced experimental results. Subsequently, we developed an improved hexapod robot that can mechanically support its weight using these legs. The robot was assembled for walking and was able to walk with a high load of 300 N.

The off-road unmanned ground vehicle (UGV) under study is made up of a body, four legs hinged to it, and a wheel at the end of each of the legs. There are rotary actuators between body and legs so that the configuration angles vary, enabling the UGV to adapt better to the terrain profile, improving its navigation capabilities. We present a numerical program based on a quasi-static half vehicle model that calculates, for a given profile, how the configuration angles must vary along the trajectory, so that the torque acting on the wheels meets some criteria, such as being the minimum or maintaining as constant as possible. Results may be helpful in control tasks, if the vehicle is equipped with a system capable of reading the terrain where it will pass, in particular when passing obstacles efficiently.

The transition from walking to running gaits in bipedal locomotion is well known from humans. One explanation for this transition is a higher energy efficiency of running gaits at higher velocities. In this paper we use a five-link planar model of a robot to investigate the transition from walking to running based on energy efficiency. For this purpose a physically motivated cost function regarding static as well as dynamic costs is introduced. Periodic walking and running gaits are generated by means of numerical optimization to find the optimal gait of a human-like model in a range from 1.5 to 2.5 m/s. At the transition velocity walking and running require the same cost. Both gaits are investigated to identify the underlying mechanisms. The computed results correspond very well to reports from biomechanics which indicates that the model is suitable for the investigation of human locomotion as well as the generation of optimal gaits for humanoid robots.

We propose a method of planning the footsteps of multi-legged robots in real-time based on A* like graph search with perceiving the surroundings by a depth sensor embedded on the robot's head. The cloud of points on the surface of the surroundings is projected on the grid map in the reference frame of the robot's head. At each step of the head legs, the planning is executed based on the graph deployed on the grid map. A series of contact points are successfully generated in realistic short time thanks to the A* like graph-search by which a suboptimal solution can be obtained. The validity is verified by computer simulation of a centipede-like robot.

This paper addresses the problem of foothold scoring for legged robots on unstructured terrain. In contrast to approaches that rely on human expert knowledge or human defined criteria to identify appropriate footholds, our method seeks to employ the robot itself as the most suitable authority to assess whether a foothold is suited for walking or not. To this end, haptic feedback from the leg/terrain interaction is applied to identify the minimal load-bearing capability of the explored foothold. Finally, the haptic feedback is associated to the foothold shape, thus allowing to estimate the robustness of footholds that are not within the kinematic range of the leg.

This paper addresses the local terrain mapping process for an autonomous robot. Building upon an onboard range measurement sensor and an existing robot pose estimation, we formulate a novel elevation mapping method from a robot-centric perspective. This formulation can explicitly handle drift of the robot pose estimation which occurs for many autonomous robots. Our mapping approach fully incorporates the distance sensor measurement uncertainties and the six-dimensional pose covariance of the robot. We introduce a computationally efficient formulation of the map fusion process, which allows for mapping a terrain at high update rates. Finally, our approach is demonstrated on a quadrupedal robot walking over obstacles.

This paper proposes the design concept that prevents its total weight from increasing. The object of the methodology is to modify a conventional robot chassis for structural simplicity by composing the chassis from an elastic material so that it can reduce vibration and shock to the body. This idea makes it unnecessary to attach additional elements, such as suspension units; in addition, it means that it is possible to easily deform the chassis by forming long air spaces, and to decrease the total weight of the robot, even when the chassis is thick enough for functionality. Structural models for analyzing the chassis flexure and stress using the finite element method (FEM) are pursued. To remove some areas from the chassis, forms of long air space are discussed and evaluated for elastic performance. The minor difference in error between the experimental and analytical data verifies that the models are useful in a practical application.

Experimental vibration dynamic characteristics of wall climbing robot with vacuum grippers and pneumatic drives are delivered. External disturbances were generated in the form of different frequencies of vibrations and vacuum pressure variation. Transient responses, frequency characteristics and phase responses are estimated. Experimental technique is discussed.

Todays walking robots are capable of walking in a wide range of terrains. One key feature, especially for exploration of unknown or extraterrestrial areas, is however rarely seen: The ability to grasp and manipulate objects or to pick up samples. In this paper we present a robust, lightweight and very versatile gripper, especially designed to be used on a walking robots leg to enable LAURON V and other robots to use their legs for manipulation tasks.

This paper analyzes second-order mobility of grasps considering contact surface geometry (metric tensor, curvature and torsion). The distance between a grasped object and each finger is formulated by using homogeneous transformation matrices. The first- and second-order partial derivatives of the distance are derived. Moreover, effects of the contact surface curvature are analyzed.

Based at CERN, the Central European Particle Physics Laboratory near Geneva in Switzerland, the Large Hadron Collider (LHC) particle accelerator has been the most powerful accelerator in the world for two decades. Utilizing the PS Booster (Proton Synchrotron) to generate and propel a 14 TeV proton beam around its 27 km circumference, the LHC provides proton collisions for experiments such as ATLAS sited 100 m underground in the beam line. In late 2012 the LHC shut down for two years for maintenance upgrades. In 2022 to extend discovery potential, the next LHC upgrade called HL-LCH (High Luminosity LHC), will rely on innovative solutions to exceptional challenges for technologies. To withstand an increase in luminosity (rate of collisions) by a factor of 10, therefore new research into radiation hard materials, structures and electronics is required. With this in mind, a new a radiation hard robot system has been developed within a new ATLAS scanning facility at the University of Birmingham Medical Physics cyclotron. This system is used to irradiate silicon sensors, optical components and mechanical structures (e.g. carbon fibre sandwiches). The robotic control system of the ATLAS scanning facility is described in detail in this paper.

The American National Standards Institute/Industrial Truck Standards Development Foundation (ANSI/ITSDF) B56.5 Safety Standard committee for safety of automated guided vehicles (AGVs) recently considered proposals for changes to improve the to make AGVs safer. The potential changes include new bumper force test methods and revisions to address sudden obstacle appearance in the AGV path. Also, the committee discussed the addition of full human form test pieces to the three current geometric obstacle test pieces. Beyond these changes to the safety standard, the National Institute of Standards and Technology (NIST) has suggested a new AGV Performance Standard be established through ASTM International to provide AGV users and manufacturers with non-safety test methods to relate measured vehicle performance to required tasks. The ASTM AGV Performance standard has been approved. Both the suggested safety standard improvements, and the proposed new performance standard, are described and/or referenced in this paper including illustrative laboratory measurement data and analysis to foster and support discussion.

In this paper, we propose a design of a new spherical mobile robot with high-dynamic and all-terrain motion. The robot has two single-axis gyroscopes on a revolvable platform for this purpose. It incorporates a jumping mechanism, as well. The structural design of the robot is ready; the movement process on a plane is modelled with and without activated gyroscopes.

A portable underwater robot system is required for observations in shallow water, because the weight and volume of a large robot and a large robotic system make them difficult to utilize under these conditions. We have designed an underwater robot with a tensile structure composed of thruster units. Utilizing these thruster units as struts, the robot had a lightweight body and occupied a large volume in space when moving. This manuscript shows the verification of this concept through geometric analyses and experimental evaluations.

This paper introduces a study of a concept of flexible crawling mechanism to design an industrial underwater cleaning robot, which is evaluated from the viewpoint of the capability to work underwater, scanning the desired surface, and bearing the reactions. This can be used as a robotic application in underwater surface cleaning and maintenance. In this study we focused on realizing the adhesion on the surface in stationary and in motion, bearing reactions, enabling the needed locomotion types for scanning, and achieving the stability in different situations on the surface.

In this paper, we proposed the unique method for climbing over a step for a five-wheeled wheelchair with an add-on single wheel drive system. The manual wheelchair with a single wheel drive system forms a five-wheeled wheelchair. This add-on system enables the wheelchair to perform a static wheelie motion in which the front casters of the wheelchair are lifted up. We propose a step climbing method by the static wheelie motion. To realize the static wheelie, kinematics model of a static wheelie motion is analyzed. By the experiments with a prototype, we confirm the wheelchair can perform a wheelie motion without slipping with small force by the add-on drive system.

In this paper, an adaptive neuro-fuzzy inference system (ANFIS) controller is designed to control knee joint during sit to stand movement through stimulation of quadriceps muscles. The developed ANFIS works as an inverse model of the constituent functional electrical stimulation (FES)-induced quadriceps, and lower leg system. It is also used for direct feed-forward control. Moreover, a proportional-integral-derivative (PID) controller is used for feedback control. The overall control strategy is designated as ANFIS-PID control method. The results demonstrate that the control approach reduces deviations between desired and actual knee joint trajectories and succeeded controls knee joint motion during sit to stand movement. Promising simulation results provide the potential for feasible clinical application in the future.

Patients use orthoses and prosthesis for the lower limbs to support and enable movements, they can not or only with difficulties perform themselves. Because traditional devices support only a limited set of movements, patients are restricted in their mobility. A possible approach to overcome such limitations is to supply the patient—via the orthosis—with situation-dependent gait models.

To achieve this, we present a method for gait recognition using model invalidation. We show that these models are capable to predict the individual patient's movements and supply the correct gait. We investigate the system's accuracy and robustness on a Knee-Ankle-Foot-Orthosis, introducing behaviour changes depending on the patient's current walking situation. We conclude that the here presented model-based support of different gaits has the power to enhance the patient's mobility.

The paper is focused on examining the possibility of increasing the effectiveness of knee rehabilitation by means of robotic systems. The case study for 16-year old patient with Ilizarov apparatus mounted on the shin is presented. A set of tests incorporating the commercial rehabilitation system and X-ray images is ascribed. Recorded data provides information concerning e.g. the force and velocity trajectories of the knee. The design of rehabilitation manipulator fulfilling the requirements is presented and its structure concerning hardware and software parts is discussed shortly.

Simulation frameworks are common tools to test new algorithms or to analyze the behavior of a robot before executing the control software on the real machine. This tremendously reduces time and effort during the development process. This paper presents a component based framework for simulating different wall-climbing robots that use negative pressure adhesion in combination with a drive system. Key aspect of the simulation is the vacuum adhesion system: Surface characteristics and features in the environment influence its overall performance, which is calculated based on a thermodynamic model of the airflows. The framework tremendously improves the development process of the new wall-climbing robot CREA by the possibility to validate controllers and algorithms offline and in realtime beforehand.

This article presents an application for dynamic simulation of legged robots based on a physics engine. In the presented application an iterative solver is supported by analytical equations of the dynamics and software modules for collision detection, environment modeling and visualization. The presented application of the simulator allows for development and verification of control algorithms before their implementation on the real robot.

This article concerns current progress in the development of multi robot simulation for TIRAMISU project. This simulator is designed for training of UGV (Unmanned Ground Vehicles) operators in cooperative mission execution. The core components of the system are implemented using VORTEX physics simulation engine with OSG (Open Scene Graph) used for rendering. The engine provides an accurate physics simulation for robots working on a single stage. The main goal during development was to prepare a multi robot architecture for the simulation. The challenge was to integrate all simulation components into a common framework, therefore allowing the robots to interact with each other, without lose of simulation accuracy. Current version of the simulator has two types of robots: a) iRobot-PacBot b)LOCSTRA - a TIRAMISU robot for humanitarian demining. An example of multi robot scenario, transportation of UXO (UneXploded Ordnance), will be discussed.

In this paper we investigate computation-efficient methods for real-time filling-in of discontinuities in terrain elevation maps by exploiting the knowledge about the terrain surrounding the gap. Two algorithms adopted from the area of computer vision are tested – originally, they were designed for inpainting of digital images, that is the technique of making undetectable modifications to images. These algorithms are adopted for filling-in the missing areas in elevation maps, and tested on synthetic and real terrain map examples.

This article focuses on the influence of walking speed and direction of the robot movement on the tactile ground classification. The perception system comprise force/torque sensor mounted on the foot of a six-legged robot. The force/torque signals are registered during the negotiation of several terrains. Next, based on the statistical or spectral analysis of the signal the robot is able to classify the terrain. In this paper we are concentrated on the influence of the walking speed and the direction of the robot movement on the classification performance. The results obtained proved that it is possible to learn the characteristics of the terrain using generalized classifier, which is trained on the dataset containing measurements acquired for the whole range of the selected gait parameter.

The automation of non-destructive testing (NDT) of large and complex geometry structures such as aircraft wings and fuselage is prohibitively expensive though automation promises to improve on manual ultrasound testing (MUT). One inexpensive way to achieve automation is by using a small wall climbing mobile robot to move a single ultrasound probe over the surface through a scanning trajectory defined by a qualified procedure. However, the problem is to guide the robot though the trajectory and know whether it has followed it accurately to confirm that the qualified procedure has been carried out. This paper describes the development of an inexpensive wireless system comprising of an optical spatial positioning system and inertial measurement unit (IMU) that relates the 3D location of an NDT probe carried by a mobile robot to a computer aided drawing (CAD) representation of the test structure in a MATLAB environment. The aim is to use this capability to develop tools to guide the robot remotely to follow a desired scanning trajectory, obtain feedback about the actual trajectory executed by the robot, know exactly where an ultrasound pulse echo was captured, map identified defects on the CAD, and relate them to the real test object.

Most of the existing path planners for traversing over rough terrains use the single-valued probabilistic properties of the terrain with the extension of considering the robot's dimensions to build the cost function. The present work proposes a path planner for a tracked mobile robot to traverse over rough terrains using the robot's tip-over stability as its cost function. The contacts that the robot makes with the terrain determine the pose of the robot and in turn its tip-over stability. The estimation of the robot's pose is formulated as a linear complementary problem (LCP) and solved using the Lemke's method. We show some examples on searching paths that optimize for various cost functions over a randomly generated rough terrain. We also validate the performance of our pose estimator by comparing their results to those obtained from a dynamic simulator (MSC Adams).

The aim of this paper is to verify whether the well known control algorithms for double inverted pendulum could be used for stabilization of the Acrobot after impact. Both methods utilize feedback linearization techniques. The first control algorithm is based on careful selection of output function that leads to a proper relative degree of the system. The second approach makes use of system's kinetic symmetry property with backstepping procedure.

A suspended robot for surface cleaning in silos is presented in this paper. The suggested concept is a reasonable compromise between the basic contradicting factors in the design: small entrance and large surface of the confined space, suspension and stabilization of the robot. A dynamic study for the suspended robot is presented in this paper. A dynamic simulation in MSC ADAMS is carried out to confirm the results from the theoretic study.

Currently most researches of the cleaning robot focus on sensor technology, location and environment modeling theory. In contrast, there are fewer researches to improve the cleaning robot's performance from the aspect of mechanism design. In this paper, we propose a new concept of a multi-function robot by using modular design methods. Compared to the existed ones, our robot is able to accomplish a variety of cleaning tasks on one platform. Lastly, we use experiments to validate the proposed robot and its functional modules.

In the aftermath of a war, the existence of buried landmines poses special threats for the local population and human deminers. Although the problem is arguably getting worse, there is a strong potential for robotic solutions to tackle it. In this paper, we outline the challenges involved in humanitarian demining and the technologies currently used in the field. Based on this, we discuss the requirements for proper robotized solutions and overview the effort that has been made by the research community and the trends that have emerged.

The aim of this work is to provide an overview on the agricultural machines adapted for humanitarian demining. It is shown that agricultural tractors have been successfully employed to perform demining operations without recurring to expensive demining machines, not affordable for countries in difficult economical conditions, such as the ones affected by landmines. The state of the art presented includes the vehicle control architecture and the sensors used on such machines.

Wall-climbing robots are of great benefit in fields of application, which are dangerous or difficult to handle for humans. But nevertheless, most of the existing systems did not exceed experimental stadium. This paper will present the new climbing robot CREA which makes a step further from a research prototype to a device which can be applied for inspection and maintenance tasks of large concrete buildings. The system uses three driven and steerable wheels for locomotion and eleven individual adhesion chambers. The paper will introduce the hardware and software components as well as aspects for controlling and sensing and show first experimental results with this robot.

The deep understanding of the human activity is an essential key for a successful Human-Robot Interaction (HRI). The translation of the sensed human behavioral signals/cues and context descriptors into an encoded human activity is still a challenge because of the complex nature of the human actions. We propose a multilayer framework for the understanding of the human activity and suitable for being implemented in a mobile robot. It is based on the ideomotor theory which argues that each human action can be seen as goal-directed movements that cause intended effects in the environment. The perception layer collects data related to the kinematics and dynamics of human body and the environment/context descriptors; the classification layer combines a segment-based Support Vector Machine (SVM) method with the Video Annotation Tool from Irvine, California (VATIC) for the classification of the elementary actions; the interpretation layer allows the understanding of goal-directed activity by using a fuzzy logic-based decisional engine (developed with SpirOps AI software). In this paper, we are limited to detail the classification method, the tools and the results.

The present paper identifies the main ethical issues the EU FP7 MOnarCH Project has to address. The mission of the MOnarCH Project is to contribute to improve the quality of life of inpatient children by having robots interacting with them in distinct contexts in a hospital environment. The paper discusses the ethics challenges in MOnarCH.

This paper proposes an approach on the design of a normative rational agent based on the Belief-Desire-Intention model. Starting from the famous BDI model, an extension of the BDI execution loop will be presented; this will address such issues as norm instantiation and norm internalization, with a particular emphasis on the problem of norm consistency. A proposal for the resolution of conflicts between newly occurring norms, on one side, and already existing norms or mental states, on the other side, will be described. While it is fairly difficult to imagine an evaluation for the proposed architecture, a challenging scenario inspired from the science-fiction literature will be used to give the reader an intuition of how the proposed approach will deal with situations of normative conflicts.

This is a shorter version of a more extended article. Please consult our full work for more details.

This paper explore the possibility of a new philosophical turn in robot-ethics, considering whether the concepts of Emanuel Levinas particularly his conception of the ‘face of the other’ can be used to understand how non-expert users interact with robots. The term ‘Robot’ comes from fiction and for non-experts and experts alike interaction with robots may be coloured by this history. This paper explores the ethics of robots (and the use of the term robot) that is based on the user seeing the robot as infinitely complex.

The work presented in this paper is part of the ongoing effort by the UK Robot Ethics (UKRE) forum, which has been formed as part of the British Standards Institute (BSI) AMT/-/2 Committee on Robots and Robotic Devices. The UKRE forum has identified a range of ethical issues related to the lifecycle of robots within the merging robotic technology sector that need close attention. These are categorized into Societal, Application, Commercial/Financial and Environmental issues [1, 2]. The range of issues within each category may vary from one country to another, and even from one region to another within the same country, depending on people's perceptions and expectations…

The following section is included:

• AUTHOR INDEX