Human-Centric Robotics

The following sections are included:

• Title
• Preface
• Conference organisers
• Conference committees and chairs
• Sponsors and co-sponsors
• Contents

When humanoid robots are operated in a human environment, multiple contacts with the environment or human are often required. In these situations, compliant motion control can be very effective in dealing with contact uncertainties. In this talk, our recent research on compliant whole-body motion control of humanoid robots will be presented. First, I would like to start with our experience of DARPA robotics challenge, where the importance of compliant motion control in multi-contact situations is motivated among many other important technologies. For example, during door and drill mission. After the competition, much effort has been made to realize compliant motion control of humanoids, using two different platforms: position-controlled robot and torque-controlled robot. I would like to present our recent results in implementing compliant motion using our humanoid robots. Future perspectives in these directions can be seen through experimental demonstrations. Finally, other exciting work in our research group will be presented in relation to humanoid technologies, such as an autonomous vehicle, robotic hand, rehabilitation device for back pain, and CPR robot.

In the future, humans and robots will have to work together in complex environments, to solve difficult problems and perform hard tasks, as a team. Thus, methodologies to enable robots to learn, interact and cooperate with their human counterparts are needed, in order to allow this joint work. This talk will be focused on methodologies developed for creating multi-robot and humanrobot heterogeneous teams with emphasis on the cooperation, interaction and learning methodologies developed on our projects: FC Portugal and HearBo. FC Portugal project developed methodologies for creating heterogeneous robotic soccer teams capable of following human/robot coach high-level advice using concepts such as strategy, tactics, formations and setplays. The project also developed methodologies for learning humanoid robot complex skills, using simulation, such as walking, kicking, getting up or passing/receiving a ball, although the methodologies developed may also be used for other types of skills and robots. The project enabled our research groups to win 4 World and 10 European robotic soccer championships of distinct leagues/competitions. At HearBo, together with Honda-RI Japan we developed human robot learning and interaction methodologies for humanoid robots that dance to the beat. The talk will be illustrated with videos from both projects.

Modern robotic technology has been an active research area for close to 50 years. From the beginning, it offered great promise to benefit society. This presentation discusses how well robotics has fulfilled its promise to improve people’s quality of life. During these 50 years, robots have been applied and/or considered for a wide range of tasks. It is concluded that robotics has yet to make the dramatic impact on the quality-of-life for many people. However, there remains great potential for this to change in the near future given the recent rapid advances being made robotic technology researchers.

Reliable Non Destructive Testing (NDT) is vital to the integrity, performance management and sustainability of capital assets in safety critical industries such as oil and gas, aerospace, transportation, power generation and off-shore and subsea operations. The talk will highlight opportunities to improve the NDT of industrial structures and decrease the cost of inspection by automating the NDT with mobile robots. The challenge is to develop robots that can provide access to test sites and perform reliable NDT on very large vertical structures or structures located in hazardous environments thereby eliminating the large expense of erecting scaffolding or lengthy preparation for rope and platform access before inspection can start. The presentation will show climbing and swimming robots developed to detect weld and corrosion defects on ship hulls, floating platforms, mooring chains, petrochemical storage tanks, pressure vessels, concrete structures, wind blades and aircraft wings and fuselage. These developments provide the possibility of saving costs by reducing outage times or (where possible) carrying out the NDT in-service thus preventing expensive outages.

In this paper, a walking pattern generation algorithm for a lower limb exoskeleton is discussed. The algorithm is designed for the case when the exoskeleton is moving on an uneven terrain. The algorithm includes footstep planning, trajectory generation for the exoskeleton’s feet and a method for finding the desired orientation of the feet. The proposed algorithms have been tested using numerical simulation, some of the results of which are shown in the paper. The advantages and limitations of the proposed algorithms are discussed.

In the growing field of powered orthotics, wearable and rehabilitation robotics the compact effective actuators remain one of the objects of numerous research and elaboration. This paper presents design and control strategies of an actuator developed for the therapeutic purposes and assistance during rehabilitation. This pilot version of an actuator consists of electrical drive, ball-screw, force and position sensing and control electronics. This paper as well describes construction and operation principle of a force measurement unit with an elastic element. Based on the developed electromechanical system we perform a force control and conduct a research on the orthosis positioning with specified impedance.

Spinal Muscular Atrophy (SMA) is the second most frequent rare disease in infancy. Life expectancy of these children is relatively short, caused by the deterioration of the respiratory function, which is accelerated by the occurrence of scoliosis. It is hypothesized that maintaining walking would significantly reduce or delay the onset of related complications, improving life expectancy and life quality. However, to date there is no device to help these children ambulate. This paper presents an overview of the ATLAS 2020 project for developing a gait-training wearable exoskeleton for SMA children. Such a device is expected to increase these children's quality of life, achieving a reduction of disability and increased functional independence. Clinical evaluation of the gait exoskeleton for gait training of SMA is described and lessons learnt are finally stated.

This paper proposes a novel robotic walker with standing assistance function. Our system focuses on domestic use for elderly people who is low level of care and need nursing in their day-to-day lives. Usually, these patients require a partial standing assistance only when they need it, not a full assistance during standing motion such as a hanging by the lift. The widely and easily use of such assistance in daily life will be successful in ensuring safety and providing an inexpensive manufacturing cost. These two opposed requirements have been realized with our developed robotic walker. Our assistive system consists of a powered walker and a standing assistance manipulator, and is small enough to be used in a typical narrow room such as a bathroom. When our proposed system assists a standing motion, it uses wheel actuators on a powered walker for stabilizing its user as well as for lifting up the user. Furthermore, our proposed system has several standing assistance patterns which are designed for the user of the different symptom and selects the suitable one for its user. The performance of our proposed system was verified through experiments using our prototype with elderly and handicapped subjects.

This paper presents the outcomes, challenges and evolution of the active pediatric orthosis ATLAS 2020 throughout the clinical tests performed in the hospital Sant Joan de Deu in Barcelona (Spain) during three months of clinical trials. The test was carried out with 7 children with spinal muscular atrophy (SMA). The test evaluated the safety and performance of the exoskeleton with both psychological and physiological measures. The test was successfully passed. All children within the inclusion criteria walked more than ten meters, comfortably and safely. During the clinical trials, both psychological and technical issues were raised. In this paper, these issues are analyzed in detail. Psychologically it is necessary to prepare the children and perform several previous tests to face the fear of using the device and show them that it is safe. From a technical point of view, it was necessary to develop algorithms for the generation of synthetic trajectories to address the patientȉs joint limitations. In addition, strength control techniques were developed to guarantee good support while guaranteeing the safety comfort of these patients that are especially sensitive. In general, it is concluded that the development of an exoskeleton should be focused on a disease in both its mechanical design and its functionality.

Human muscle fatigue is identified as one of the causes to musculuskeletal disorder (MSD). The objective of this paper is to investigate the effect of an exoskeleton in dealing with muscle fatigue in a virtual environment. The focus of this work is, for the exoskeleton to provide support as needed by human joint. A (Proportional, Integration and Derivative) controller is used for both human and exoskeleton. Simmechanics and Simulink are used to evaluate the performance of the exoskeleton. Experiments show that human is able to maintain doing the job by wearing the exoskeleton.

The bio-inspired robotics use functional elements of natures for inspiration. The development of the TB-Horse II prototype is the main target of this work. It is a bio-inspired quadruped robot with biological features in horse of the breed Mangalarga Marchador. In future, the robot can be used to rescue injured people, to carry fragile loads, among others applications. With the study of horse biodynamic, it was possible to propose the TB-Horse II. The gait marcha was implemented and validated using the Virtual Robot Experimentation Platform (V-Rep). Finally, the robot prototype was developed and the experimental validation was realized on a flat ground without obstacles.

In our previous study, we developed a lunar subsurface excavation robot that mimicked the peristaltic crawling movement of an earthworm. The robot excavated soil by using an earth auger and propelled in a hole by using peristaltic crawling. however, the earth auger could not be positioned properly. Consequently, the diameter of the hole took not constant value and the robot’s ability to propel was reduced. In this paper, to overcome this problem, we have developed a both-end supported earth auger and demonstrated its performance by comparing it with existing earth augers.

In this study, a peristaltic crawling robot that is capable of traveling over 100 m in a 100A pipe was developed. The performance of new robot was evaluated by conducting experiments in straight horizontal and vertical pipes, a 90° large curved pipe, and a 90° elbow pipe, as well as in a real sewer pipe line. A running experiment with a pressure feed pipe was performed and its usefulness was verified. Currently existing robot (PEW - RO) has low robustness and running speed in a real environment which is insufficient for practical applications. Herein, based on previous research, a new peristaltic crawling robot (PEW – RO Ad) is developed and a running test in a long distance pipe was conducted. The obtained results are compared with the running performance and robustness of the currently existing robots.

Walking robots are designed to cope with difficult, challenging terrain. Therefore, developers have to take special care not to damage or break their robots legs in such environments. The shoulder joints are especially suffering from high stress values. This work presents a concept for a robust, modular shoulder design for bio-inspired hexapods. It was evaluated and is being used for the current version of the walking robot LAURON V.

Biological systems often combine cues from two different sensory modalities to execute goal-oriented sensorimotor tasks, which otherwise cannot be accurately executed with either sensory stream in isolation. When auditory cues alone are not sufficient to accurately localise an audio-visual target by orienting towards it, visual cues can complement their auditory counterparts and improve localisation accuracy. We present a multisensory goal-oriented locomotion control architecture that uses visual feedback to adaptively improve acoustomotor orientation response of the hexapod robot AMOS II. The robot is tasked with localising an audio-visual target by turning towards it. The architecture extracts sound direction information with a model of the peripheral auditory system of lizards to modulate locomotion control parameters driving the turning behaviour. The visual information adaptively changes the strength of the acoustomotor coupling to adjust turning speed of the robot. Our experiments demonstrate improved orientation towards the audio-visual target emitting a tone of frequency 2.2 kHz located at an angular offset of 45° from the robot.

This paper describes a compact climbing robot called Cubic-Climber which is driven by two differentially driven magnetic wheels. It is intended for the inspection of the inner surface of confined steel structures which are commonly found in oil rig. It can also be used to carry a welding torch to weld the inner seams of these structures. The Cubic-Climber can turn on the spot and achieve surface transition in the confined steel structures. The magnetic wheel uses wheel-parallel-in-wheel (WpW) structure to achieve surface transition function. The contribution of this paper is the optimization of WpW in terms of the overall weight and required peak motor power through searching the Pareto frontier by genetic algorithm. The experiment results show that the proposed robot is able to achieve a smoother surface transition compared with conventional magnetic wheels.

The paper describes the design of an autonomous wall climbing robot, its control system and stability analysis. Robot sticks to the surface by use of an aerodynamic suction fan.

This paper presents the analysis of locomotion of wave-like robot over flexible surfaces. The purpose of this research is to develop and improve the locomotion of robotic probes in highly flexible environments such as biological vessels for medical purposes, especially for the gastroenterology field. This unique robot, which was inspired by nature mainly from the motion of worms, snakes, and centipedes, has a minimalistic mechanical design allowing for miniaturization. In this work, we present a numerical model of the robot interface with the surface and the results of experiments that we conducted with the robot over highly flexible surfaces.

In this paper, we analyze the locomotion of a double screw-like robot over compliant surfaces. We first develop an analytical model of locomotion over rigid surface. Then we use this model to determine the forces acting on the robot and environment to define the surface properties that allow the robot locomotion. Finally, we performed experiments using the robot to determine its thrust force and its speed over hollow silicone rubber.

This paper discusses heterogeneity on distributed pattern formation by autonomous climbing agents with simple object stacking abilities. This work was motivated by social behaviors of termite colonies, in which often build elaborate three-dimensional structures (nest towers). This paper challenges to understand the mechanism in the wonderful performance; in particular, the effect of heterogeneity on distributed pattern formation by termite-like agents, from computational and minimalistic approach. We introduce a cellular automata (i.e., spatially discretized) model, where each agent (robot) selects its action from move/load/unload based on the state of its neighboring cells. The two types of agents include what we call architect and demolish agent, and we perform analyses to show that moderate existence of demolish agents would lead to control the size of buildings.

This paper describes the development of SMORA (Servo Motor Optimised for Robotic Applications), which consists of custom hardware and firmware that includes a microcontroller and a series of sensors, allowing for the motor current, temperature and voltage to be measured in real-time as well as precise position feedback thanks to the integrated hall-effect magnetic position encoder. It also incorporates an accelerometer and a gyroscope to measure the servo body relative position and rotation.

The BSc in Systems Engineering, taught at the School of Engineering of the Porto Polytechnic (ISEP–PPorto), includes the development of team projects at the end of each semester. These projects are intended to be multidisciplinary, to allow students getting a systems perspective of the problems, and as a way to integrate knowledge from different scientific areas. To integrate the knowledge in electronics, microcontrollers and programming, common in several distinct systems nowadays, there is a course named Systems Labs I (LSIS1). This paper discusses the experience of teaching this course using a robotics based approach. The analysis is based on a survey conducted to the students enrolled in the course in the last three academic years, allowing to conclude that the overall students experience with the attendance of LSIS1 course is good.

The Portuguese Micromouse Contest^® is an innovated robotic contest in Portugal that address the need to enhance student interest and performance in science, technology, engineering, and mathematics (STEM) courses, while fostering skills that are important prerequisites for IT careers. To facilitate the access to younger students a robot kit and an Arduino library was developed and made available on GitHub. Some of the library examples include the common algorithms used to solve mazes, like the Flood Fill, Right or Left Wall Following to name a few. Recursive Backtracking algorithm is often used to solve mazes especially if there is a lot of computer power. This is not the case of the Arduino platform that only has 2.5 KB of SRAM and a clock speed of 16 MHz, so implementing Recursive Backtracking becomes quite difficult. In this work we show how to implement Recursive Backtracking to find if there are better solutions for solving the maze. We also show the memory measurements for three Arduino common platforms the Leonardo, the UNO and the DUE.

Thousands of elementary school students participate in Junior Botball Challenge exercises every year. These challenges require students to write programs for their robots and to supplement their basic robots with effectors to carry out the challenge task. This paper presents data gathered from some of the schools that have participated, with a focus on those that did NOT select students based on their interest or ability. It shows that a large percentage of typical elementary school students are able to write working C programs (when given appropriate instruction) that exhibit sequential steps and timing.

This paper describes the design and implementation of a ground-related odometry sensor suitable for micro aerial vehicles. The sensor is based on a ground-facing camera and a single-board Linux-based embedded computer with a multimedia System on a Chip (SoC). The SoC features a hardware video encoder which is used to estimate the optical flow online. The optical flow is then used in combination with a distance sensor to estimate the vehicle’s velocity. The proposed sensor is compared to a similar existing solution and evaluated in both indoor and outdoor environments.

MBZIRC is an important robotic competition that was held for the first time in Abu Dhabi in March 2017. The main objective of Challenge 1 was the autonomous landing of an UAV on a moving vehicle. This paper shows an overview of the architecture of the system and of the different modules implemented, developed by the team of the University of Catania, that achieved the fourth place among the 26 participants.

Sliding mode observer (SMO) as a nonlinear and robust observer, is believed to be able to provide all required states information for control process of quadcopter UAVs. In this paper, a comparative assessment through numerical simulation is conducted between SMO and Extended Kalman Filter (EKF) to demonstrate the performance of both estimators. The results obtained demonstrate good performance of SMO in dealing with noise and uncertainty. Furthermore, experiments are carried out to validate the performance of SMO in real-time. The results show that estimated states can track true states fast with small estimation steady state errors and the observer estimates the unmeasured states smoothly.

3D path planning with unmanned aerial vehicles in search and rescue scenarios is an important research area, due to the ability to explore damage areas that could be inaccessible for vehicles like ground robots. This paper presents two innovative real-time path planning algorithms based on PRM (Probabilistic Road Map) able to be implemented in UAV’s denoted by Grid Path Planning Roadmap Planning (GPRM) and the Particle Probabilistic Roadmap (PPRM). With the requirement of being implemented in a real search and rescue scenario like the EuRathlon competition, the GPRM method will produce a roadmap building step with obstacles inside a predefined grid while PPRM will follow a different approach by introducing an associated probability to each computed path in order to support the next sampling step path planning iteration. Both methods were evaluated and compared with the well known 3D path planning PRM in a search and rescue earthquake simulation environment developed in MORSE (Modular Open Robots Simulation Engine).

This paper describes the process of designing a reliable wall-climbing robot that can be produced easily. We found that designers can choose out of many possible technologies to make wall climbing robots. After exploration and prototyping with gecko tape, suction caps, microsplines, magnetism, impellers and propellers we came up with GECO. GECO is a wall climbing robot which uses suction to move across walls. GECO moves over different kinds of surfaces, is wireless controlled and relatively easy to build. We envision that wall-climbing robots have potential applications for example to inspect surfaces, clean and maintain buildings and many more.

This paper proposes a glass-wall climbing robot that is intended for cleaning curtain glass wall. It is based on passive suction cups mechanism. The main advantages of the proposed robot are relatively small thickness (which can allow greater payload with a given set of suction cups), easy attachment and detachment, and lower energy consumption. The proposed robot consists of a newly designed guiding rail and eight suction cup systems. The guiding rail has small thickness that results in a portable and compact design. The payload is up to 2 kg. Because the passive suction cups do not need energy input to attach to the glass wall, the proposed robot consumes relatively low energy. A prototype of the robot has been fabricated for evaluation.

The paper presents some results of the study of behavior of wall climbing robots with vacuum contact devices under water conditions. Special attention is concentrated on mechanics of vacuum contact devices equipped with “gas-water” ejector where air is supply and water is intended for vacuuming. The features of such kind of robot are analyzed. Some design peculiarities and demands for testing are adjusted. The main hydromechanical characteristics are illustrated the functional possibilities of wall climbing robots working under water.

Recent developments in the field of multi-legged mobile manipulators encouraged researcher to develop new robotic platforms combining the locomotion stability of multi-legged robots with manipulation capabilities. This paper provides an extensive bibliography that can be of help to researchers interested in further studying and designing centaur-like robots. A centaur is a mythological creature with the upper body of a human and lower body of a horse. Similar to the centaur, half-human and half-horse composition in robotics can be exploited to achieve stability and manipulation capabilities.

Wall climbing robots are extremely useful for deployment in hazardous environments and operation at great heights and other hard to reach space. However, these robots usually require large and heavy suction pads and associated cylinders for supporting the payload and effectual movement during climbing and inspection process and are very difficult to miniaturise. A new type of robot mechanism has been developed where movements are powered by electric motors and hence can be made on a much smaller scale and with the promise of high speed of operation.

In this study, we developed a wall-climbing robot for aircraft body inspections. To inspect aircraft body, a robot is required to adhere to the curved surface of the aircraft made of nonmagnetic material and move quickly and stably. Therefore, we equipped the developed robot with omniwheels for movement and negative pressure suction. We then fabricated the developed robot and examined whether the robot can move on aircrafts.

There are many man-made structures near the ocean, in the so called splash zone. These structures are submitted to corrosion and need to be inspected periodically, which is difficult to be performed by humans. Therefore, automated solutions should be devised, able to withstand the conditions found there. Given that some animals live in this environment, the authors propose the development of a biological inspired robot for achieving such inspection tasks. With this purpose, a biomechanical study of the spider crab was developed, focusing on the anatomy and locomotion of this animal, using the Matlab/Simulink SimMechanics toolbox.

A variety of automated solutions for non-destructive testing have been developed to facilitate in-situ sensing for inspection requirements to ensure safety and effective operation in a variety of hazardous application scenarios. The paper focuses on climbing robot technologies which have been developed to monitor the condition of assets such as concrete pillars, highway bridges, tunnels, wind turbine blades and towers, hull ships, pipelines, nuclear installations, etc. A review of the robot technologies is presented together with a number of climbing robots being developed via magnetic and vacuum adhesion.

The paper presents an overview of pipe inspection applications and describes a number of robotized solutions which have been realized for hazardous environments. Pipelines are valuable and important assets vital for moving combustible products in various installations; it is vital that reliable and competitive means of transporting the various oil and gas products and by-products are needed. The pipeline assets have long and active lives and hence the pipe networks have to be continuously monitored for ensuring safety and effective transportation. A variety of solutions such as intelligent pigging robots, crawling robots as well as free swimming robots have been developed. Few exemplar pipe inspection robotic systems that have been developed in R&D projects are described.

Petrochemical storage tanks are inspected mostly with outages to assess the extent of underside corrosion on the tank floor. Emptying, cleaning and opening a tank for inspection takes many months and is very expensive. Inspection costs can be reduced significantly by inserting robots through manholes on the tank roof and perform nondestructive testing. The challenge is to develop robots that can operate safely in explosive and hazardous environments and measure the thickness of floor plates using ultrasound sensors. This paper reports on the development of a small and inexpensive prototype robot (NDTBOT) which is easy to make intrinsically safe for zone zero operation. It hops around a floor to make measurements without using external moving parts. The paper describes the design, experimental testing of the NDTBOT and results of steel plate thickness measurement while operating in water.

This paper presents the architecture of a localization system for affordable service robots. The system utilizes a RGB-D sensor for visual global localization based on artificial landmarks. The unobtrusive and cheap landmarks are based on QR-codes. We demonstrate how to recognize printed QR-codes over a wide range of viewing ranges and angles, and compute the robot pose uncertainty related to the landmark measurement. The localization method is integrated in the Robot Operating System and demonstrated in a home-like environment.

This paper introduces a new terrain classification method based on visual data from a typical RGB camera and a 2-D laser scanner. The aim of the method is to robustly recognize the terrain type in front of a mobile robot that navigates in an outdoor urban environment, which consists of areas covered by specific surfaces, such as asphalt roads, pavements, lawn areas, etc. We perform terrain classification on relatively simple features based on color in the RGB images and intensity in the laser scanner readouts. Random Trees are applied as the classifier with context-aware classification using Probabilistic Graphical Model. The information about terrain type is used in motion planning.

Autonomous navigation of ground vehicles in rugged outdoor environments is still an open issue in robotics research, although it finds a wide variety of applications, from search and rescue to planetary exploration, and advanced agriculture. A common approach to face this problem is based on so-called terrain traversability analysis, which consists of assessing the difficulty encountered by the ground vehicle while moving through a certain area. This paper describes a general strategy of cooperation between a UAV and a UGV for the autonomous navigation of the latter within an outdoor zone. The aerial vehicle performs a survey of the region of interest, from which the terrain surface model is built; then a terrain traversability analysis is performed, taking into account the specific navigation features of the considered ground vehicle.

In our previous work, on a passive-based biped walker with a torso, the bifurcation gaits whose initial states are far from the unstable fixed point are successfully suppressed by our MPC-based suppression method, which is implemented only by small perturbations on the parameters. It is natural for us to think that such an idea can be transplanted to help the biped walker reject disturbances. Small perturbations on parameters mildly affect on the dynamics of the walker, so rather than promoting the inherent maximum disturbance rejecting magnitude, accelerating the convergence after being disturbed is a more reasonable and promising way to use this MPC-based method. The performance of this method is examined by a single stair disturbance and an impulsive push disturbance. The results of the simulation experiments show that the MPC-based method significantly accelerates the convergence of the gait to the cyclic pattern after being disturbed, while contributes a bit to the maximum disturbance rejection magnitude as well.

This paper proposes a bipedal locomotion control with different walking speeds in the B4LC system. Inspired by the biomechanical and biological researches, we adapt the motor patterns and reflexes at the hip, the knee and the ankle joints to extend the speed control capability. The parameters of these control units are autonomously searched using the Particle Swarm Optimization method and the parameter functions with respect to the walking speeds are further constructed. By validating on an anthropomorphic simulated biped, the experiment results show significant improvements on bipedal speed control from 0.63m/s to 1.63m/s.

In this paper, we propose the gait control strategy for a six-legged robot walking on rough terrain. To walk efficiently on rough terrain the robot uses proprioceptive sensors only. The robot detects contact with the ground and uses Attitude and Heading Reference System (AHRS) unit to measure the inclination of the platform. We propose a single-step procedure to compute inclination of the robot’s platform taking into account the terrain slope and kinematic margin of each robot’s leg. Additionally, we use a procedure, which keeps the robot stable during walking on rough terrain. We show in the experiments that the robot is capable of climbing slopes inclined by 25° and walking efficiently on rough terrain.

The study of passive dynamic walking, among other techniques, is used to understand how animals walk. In our previous study, a quadruped robot inspired by passive dynamic walking was developed. This robot (Duke-II) performed quasi-passive dynamic walking, which was inspired by passive dynamic walking and intended to achieve level walking by only applying a small amount of energy. In this paper, we focus on the influence of the trunk structure on walking ability. We developed a new trunk part for Duke-II and investigated the influence of the trunk structure and the vibration characteristics of the robot on its gait and walking speed.

Myriapod locomotion has an advantage over wheeled and tracked vehicles on a rough terrain, as each leg can discretely contact the ground at several points. However, there are many unanswered questions regarding the mechanism for myriapod locomotion, particularly with respect to manner of legs movement and torso undulation. The typical myriapod robots, however, were originally large and heavy in order to actuate numerous joints; thus, it is difficult to believe that these robots are able to synthesize aspects of intelligence, such as adaptability, of Myriapoda. Therefore, the aim of our study is to develop a light, simple, and adaptive myriapod robot based on passive dynamics. We assume that interaction between the leg and environment includes an implicit control law, which enhances mobility and stabilizes locomotion. The mechanical aspects of the torso and legs such as flexibility may be the basis of the implicit control law. Thus, in this study, we develop a novel prototype of the myriapod robot called i-CentiPot 01 by implementing passive dynamics, and subsequently analyze its locomotion and conduct some field test in order to demonstrate its adaptability in accordance with the implicit control law given by the passive dynamics.

The motion of the upper body of bipedal locomotion, which includes the trunk and arms, influences the characteristics of the bipedal locomotion such as stability and velocity. This study focuses on a viscoelastic trunk mechanism with swinging arms, and observes the effect of the trunk and arms on bipedal locomotion. We develop a physical robot and a simulation model, and we observe the robot and the model while each walks with swinging arms. The results of physical experiment show that the arm swinging with the viscoelastic trunk mechanism suppresses the yawing on the supporting foot. The simulation results show that the arm swinging with the viscoelastic trunk mechanism allows greater distance locomotion. These results indicate that the viscoelastic trunk mechanism with swinging arms has an advantage over the rigid trunk without swinging arms.

Multi-legged robots are expected to be able to function in extreme environments to perform tasks such as planetary exploration and nuclear power plant maintenance. Localization of the robot body is important in the walking directional control of six-legged robots. Currently, localization measurements are carried out by extracting two landmark points using a laser range finder. However, it is not possible to estimate the exact amount of movement using this method. Thus, the authors propose the application of the iterative closest point algorithm (ICP) as a localization method in this study. The ICP algorithm is a well-known method of registering two shapes composed of point clouds to estimate the rigid transformation that minimizes the distance between corresponding points in the two point clouds. By iteratively obtaining the correspondence and estimating the rigid transformation, registration for given shapes can be performed. Moreover, the effectiveness of the walking directional control of a six-legged robot performed by localization using the ICP algorithm was verified in this study.

According to the similarity in kinematic architecture, bipedal robots are the most appropriate type of robot to operate in humanoid environments. Most of the humanoid robots have more than 20 degrees of freedom (DoF), therefore they have complex kinematics and dynamics. Due to these complexities, developing a stable walking engine is a difficult subject which is still one of the main challenges. In this paper, a hierarchical walking engine is presented which tries to fade the complexities and increases the flexibility and portability. To generate the reference trajectories of walking, Linear Inverted Pendulum Plus Flywheel Model is used. We enhanced this model to release the height constraint of the Center of Mass (CoM). This enhancement not only provides more natural motion but also it provides larger stride.

The reliability of the proposed structure is verified through real experiments for an 110cm bipedal robot. The experimental results show the performance of this controller to keep robot’s stability during walking. The average speed of walking that we have achieved was 20cm/sec.

In this paper we describe the steps that allowed us to realize real outdoor experiments of HyQ bounding at different speeds and performing omni-directional maneuvers. The strategy is composed of two parts: the first one is an offline optimization that finds a stable periodic limit cycle which represents the baseline bounding gait; the second part is a speed controller that adjusts online the main gait parameters based on the high-level speed commands coming from the external operator. In the tests HyQ reached a forward speed of 2.5m/s, lateral speed of 1m/s and angular speed of 50deg/s in simulation and respectively 1m/s, 0.5m/s and 30deg/s on the hardware experiments.

This paper illustrates the results of a validation procedure for the computer simulations of our quadruped robot HyQ. We show how simulated and real data recorded during locomotion tests are substantially consistent, providing an argument for the reliability of our simulation software. The main contribution of this work is to illustrate a basic yet effective software system that allows us to simulate – and also control in the real world – a complex articulated robot that can walk and run.

Several actuators to be implemented in bipedal robots have designed since several years up to the present with different characteristics and objectives. In this work a nonlinear variable transmission actuator for biped robots is presented. This special actuator is embedded in some joints of the lower extremities of the biped robot in order to improve the locomotion characteristics, specifically, in the foot-ground contact. The idea is the interaction of the foot with the soil tends to be more compliant. Since a force sensor has been installed on one of the elements of the nonlinear variable transmission actuator has been possible to implement force feedback control and, consequently, some experimental results have been obtained.

In this paper, we apply decentralized feedback control for a tripedal walking robot with three radial legs driven by linear actuators. The control scheme, originally proposed by Owaki and Ishiguro et. al., consists of a phase oscillator and sensory feedback of reaction force from the ground, where the control law for each leg is decoupled from the others (i.e., it has no explicit feedback of the other legs’ information). We show that rotary and forwarding locomotion successfully emerge using the control method, depending on the choice of frequency ratio of the oscillators.

Correction control is analyzed for adaptation of robots contact vacuum devices to unknown in advance surface quality. Two kinds of feedback loops are under consideration – pressure supply valve control and force control of the robot’s leg with pneumatic vacuum contact device. Simulation models were developed to study correction control. The results of simulation are suggested to improve essential dynamics of wall climbing motion over non predicted environment.

The paper deals with modeling of a thermomechanical actuator for a walking microrobot. The actuator is modeled by a linkage of rigid bars connected by elastic joints. The configuration of the linkage is studied as a function of the temperature and applied loading force.

Non-destructive evaluation of structures such as ships and large oil tanks can be rapidly carried out using robots employing magnetic adhesion. This eliminates the need for the use of laborious suction systems, usually employed in climbing robots. However, since the operation of the robot is very much dependent on the speed at which the robot can move before the adhesion is lost, there is a need for determining limits at which magnetic force becomes ineffective. Finite element models and dynamic solution of magnetically supported robots have been presented in order to determine the speed limits at which they can operate.

The focus of this article is the design and development of a novel multipurpose omnidirectional modular all-terrain mobile robot. The locomotion is achieved by means of multiple screws and can bear a wider variety of terrain than ordinary robot using a logical combination of the angular speeds of each screw. The kinematic model of the proposed robot is described in detail and CAD models were generated. Simulations of a trajectory tracking control were performed to test the proposed robotic architecture.

Traditional methods for robotic biped locomotion employing stiff actuation display low energy efficiency and high sensitivity to disturbances. Legged locomotion can be modelled as an hybrid system, where continuous dynamic flows, such as the single or double support stages, are interrupted by discrete jumps, such as heelstrike or lift-off. Traditional control systems are not suited to deal with hybrid systems or with the compliance added by passive elements. A Model Predictive Control (MPC) approach is proposed to deal with the hybrid system dynamics. The controller generates energy efficient gaits for a simulated Simplest Walker (SW) mechanism, tracks the gait trajectories in the presence of sensor noise and small disturbances and is able to adapt to strong and impulsive pushes.

Force and tactile sensing is required for robots interacting autonomously with their environment. Unfortunately, most force sensors available today are still too expensive to be deployed on a large scale. In this paper we introduce a modular approach to design and integrate low-cost force and tactile sensors directly into 3D-printed robot parts. Based on commodity optical proximity sensors embedded into deformable cantilever structures, sensitivity and load capacity can be selected in a wide range. Our modular CAD-library allows the designer to interactively dimension and shape the sensor for a given purpose.

The motion trajectory planning method for quadruped robot with static gait is introduced in this paper. Firstly, the leg swing order is designed to keep good stability, as well as the single leg swing and four leg stance phase in a gait period. Then, the minimum longitudinal stability margin is calculated, which is between the edge of the triangle support polygon to the projection of the center of the robot mass in the support plane when one leg is swinging. Based on the calculated minimum margin, the common stability polygon during the process that the two legs on the same side swing is got, and the center of gravity in this polygon is defined as the target position of the center of body mass motion in the end of the four leg stance phase. With the above results, the body motion trajectory is calculated with the sine wave function. Next, the foot swing trajectory is fitted by the three spline curve, which can avoid and cross some unknown obstacles. Finally, the simulation results show that, with the designed trajectory planning method, the quadruped robot can walk with good stability and continuity in the static gait.

The article deals with control of a robot with three omni-wheels. The feature of this robot is the triangular platform with a right angle. The steering function is of special interest of the paper. The explicit formulae of moments applied to the wheels are obtained for the robot’s movement along a specified trajectory for two particular cases.

Scheduling assumes a crucial importance in manufacturing systems, optimizing the allocation of operations to the right resources at the most appropriate time. Particularly in the Flexible Manufacturing System (FMS) topology, where the combination of possibilities for this association exponential increases, the scheduling task is even more critical. This paper presents a heuristic scheduling method based on genetic algorithm for a robotic-centric FMS. Real experiments show the effectiveness of the proposed algorithm, ensuring a reliable and optimized scheduling process.

In conventional research methodologies, multiple cameras are installed on a flexible mono-tread mobile track in order to teleoperate the device using the images displayed on the cameras However, it is desirable to have as few cameras as possible, owing to the limited space available on the device to place the cameras and the need to collectively align the optical axis of the cameras. This paper describes an experiment wherein we operated an experimental device using one installed camera to supplement the entire area covered in the field of view by combining the original and past images.

In this paper, we propose the teleoperation interface for a hexapod walking robot. The interface is based on the Kinect sensor and hand tracking libraries. The set of gestures is used to switch between motion modes of the robot. The estimated position of the operator’s hands is used to define the reference motion of the robot. We show in the experiments that the operator can control the robot to reach the goal position and manipulates objects using hand gestures only.

This paper introduces an open-source software toolkit for combining the Robot Operating System (ROS) and Unity3D to versatile robotic applications involving virtual environments. Despite the availability of high-quality robot control and simulation systems like ROS, there is still no framework available for designing complex human-robot interaction tasks and making use of Virtual Reality without expert knowledge in robotics and ROS. Virtual Reality, especially involving head-mounted displays as well as various input and feedback devices can increase the experienced sensation, improve the understanding of certain 3D scenes or provide a test environment for the training of non-expert operators. As a solution, we propose Unity3D for designing the interaction interface to virtual and real robots. Our implementation of the bilateral communication layer between Unity3D and the ROSbridge and the importer for XML-based files in Unified Robot Description Format allows synchronizing the state of the real robot and the one of its simulated counterpart in the Unity3D environment in a very comfortable way (see Fig. 1). Using the proposed system we present different use cases and demonstrate how to prototype different interaction concepts efficiently.

Surface electromyography (EMG) signals classification is currently applied in various prostheses and arm controls using various classification methods. The limited robustness in practical EMG control applications has become an important matter of research consideration. The precision of EMG signal features and parameters proportionally vary with muscle fatigue (MF). The major challenge for the study is to identify the MF manifestation in the EMG signal, so that the control performance is improved. This can be done by the improvement of data collection practicality, features extraction and classification. Hence, fundamental study is performed by investigating the signals acquired from the human upper forearm (UFA) to determine muscle characteristics and to establish the inter-relationship between both muscles of the forearm and upper arm. The aim of the present study is to investigate the applicability of human UFA muscles and MF indices at various force levels of maximum voluntary contraction (MVC). EMG signals are recorded from nine (9) normally limbed subjects. The frequency domain power spectrum density (PSD) is computed in order to derive the useful characteristics of the signal. The results show that only few muscles contributes for the movement. Further analysis show that flexor digitorum superficialis (FDS), flexor carpi radialis (FCR), extensor carpi radialis longus (ECRL), extensor digitorum communis (EDC) and biceps/triceps brachii show interesting results.

We developed a flexible propulsion unit capable of three-dimensional movement. This could be used to drive a seabed excavation robot for sub-seafloor exploration. The propulsion unit comprised two passive joints and three propulsion subunits, and included a gripping mechanism and a pneumatic cylinder. We experimentally tested the performance of the propulsion subunit, and confirmed that it was able to propel itself through both a straight borehole and a curved duct, at similar speeds.

In this paper, we present two versions of a novel volume-expanding mechanism for ships and underwater vehicles to obtain large buoyancy change, and report our experimental results. Our aim is to build a buoyancy-changing device utilizing the volume change of material (paraffin wax) induced by the phase change between its solid and liquid states inspired by a hypothesis on the buoyancy adjustment method of sperm whales. The original value of paraffin wax’s volume change is too small to alter the total buoyancy of underwater vehicles. Therefore, a mechanism that creates a much larger difference in volume is required. We propose a volume-expanding mechanism with metal bellows and rods that utilizes the difference of the cross sections between the bellows and the rods. Our experimental data confirms that the proposed mechanism expands the original buoyancy change of paraffin wax by a maximum of 50 times.

Some results of underwater tests of subsea walking unit MAK-1 are discussed. During experiments in underwater conditions walking unit performance was checked and influence of walking mover design features on his dynamics, maneuverability, traction characteristics, passableness has been investigated. Also, certain attention was given to testing of methods of standalone movement control of subsea unit. Tests have shown that walking movers in subsea conditions can provide higher traction properties and passableness, in comparison with wheeled and tracked ones. Also, underwater test showed, that underwater walking device MAK-1 with cyclic type of mover surpasses known counterparts with adaptive control on speed and maneuverability. The results can be used for development of underwater walking machines and robots, intended for underwater technical works, for new industrial technologies of exploration of seabed resources and for providing of ecological safety of underwater objects infrastructure.

Mooring systems experience high tidal waves, storms and harsh environmental conditions. Therefore, ensuring the integrity of mooring chain is important. The aim of the work reported in this paper is to develop a robotic system that performs in-service non-destructive testing of mooring chains. The inspection system is an autonomous device that operates in air as well as underwater. The permanent magnet adhesion crawler robot developed can climb mooring chains at a speed of 42cm/minute with a pay load of 50N. FEA study of the magnetic adhesion module, structural analysis, prototyping and testing of the robot is presented in this paper.

The COST Action CA16116 – “Wearable Robots for Augmentation, Assistance or Substitution of Human Motor Functions” was recently started. The goal of such an Action is to create a European network in support of the academic, technological, and societal development of the field “Wearable Robots” across Europe. This paper will introduce the basics of the Action as well as clarify how interested researchers and other stakeholder can get involved.

The aim of this paper is to present the design and development of an active orthosis system (AOS) for assistance and rehabilitation. Active Orthosis (exoskeleton) is assistive device with a wearable structure, corresponding to the natural motions of the human. The exoskeleton structure includes left and right upper limb, left and right lower limb and central exoskeleton structure for human torso and waist and provides support, balance, and control of different segments of the body. The device was fabricated with light materials and powered by pneumatic artificial muscles that provide more than fifteen degrees of freedom for the different joints. The actuation of the AOS is inspired by the biological musculoskeletal system of human upper and lower limbs, and mimics the muscle-tendon-ligament structure. The system can operate in three modes - Assistive Mode; Haptic and rehabilitation device and Motion tracking system with data exchange with virtual reality. The AOS can be used for human interaction with virtual environments where motion tracking and force feedback are required. The system would be of great importance to people with limited mobility for assistive and rehabilitation tasks

Wearable robotic exoskeletons could provide a solution to the ever growing pressure on rehabilitation services. They could provide the increased intensity and required repetition giving objective feedback on patient’s progress. They can also support independent practice freeing up clinicians time resulting in improved efficiency. Integration of wearable robotic exoskeletons for rehabilitation can result in efficient and effective outcomes for all users (i.e. patients, clinicians, carers, community therapists). However, there is evidence of varied, slow or even lack of adoption and integration of these technologies into mainstream practice and in clinical pathways for rehabilitation. It is important to bridge the gap between technology developers and technology users or adopters to ensure the promotion of patient centered functional recovery. We present the results of a study aiming to compile a set of requirements specifications based on users’ interviews and surveys. We have interviewed over seventy patients and clinicians to explore the requirements of a wearable upper limb exoskeleton device defined by the users to facilitate the adoption of the technology within rehabilitation services in hospitals and in the community and ensure the efficacy of the outcome.

Patients suffering from Spinal Cord Injury (SCI) can partially restore their walking function over ground and reduce of the secondary pathologies thanks to a frequent use of powered exoskeletons for robot-assisted gait. However, many of these devices require crutches to stabilize the body or to initiate the step. This could induce high loads on the shoulder joints, leading to shoulder pain and pathologies in SCI patients. This study presents a methodology to evaluate shoulder joints loads during robotic assisted gait, which was applied to an expert user of a Rewalk® exoskeleton. A pair of wireless instrumented crutches, designed by the authors[1], measure the crutches’ forces and movements. A simplified mechanical model of the patient is then simulated in OpenSim to perform both inverse kinematic and inverse dynamic analysis. Such a system could be applied during the initial training of new users, to help guide both patient and therapist towards an optimal usage of the exoskeleton.

Empowering robotic solutions are exploited for industrial applications in order to reduce/limit risk factors related to musculoskeletal disorders (MSDs) and to improve the capabilities of humans. This paper aims at proposing both (a) the design and (b) the control of a cooperative manipulator in order to empower humans in onerous industrial tasks execution. (a) a dual driven actuation (DDA) is proposed and described for the shoulder joint of an empowering robotic system. (b) a fuzzy impedance control based approach to assist the human operator while lifting (partially) unknown weight components is proposed. The control method has been validated in a hatrack-like component installation, case-study related to the H2020 CleanSky 2 EURECA project. The proposed application has been shown during the KUKA Innovation Award. As a test platform, a KUKA iiwa 14 R820 has been used.

Real-time locomotion mode recognition can potentially be applied in the gait analysis as a diagnostic tool or a strategy to control the robotic motion. This research aimed the development of an automatic, accurate and time-effective tool to recognize, in real-time, the locomotion mode that is being performed by a humanoid robot. The proposed strategy should also be general to different walkers and walking conditions. For these purposes, we designed a strategy to identify, in an offline phase, the suitable features and classification models for the real-time recognition. We explored several classification models based on two machine learning approaches using the features previously selected by principal component analysis and genetic algorithm (GA). The validation was carried out for distinct walking directions and speeds of DARwIn-OP. The offline analysis suggests that the most skilled models are the ones created by weighted k-nearest neighbors (KNN), fine KNN, and cubic support vector machine using 2 features selected by GA. Results from the real-time implementation highlight that weighted KNN exhibits a higher recognition performance (accuracy > 99.15%) and a lower elapsed time in the recognition process (89 ms) comparatively to the state-of-the-art. The proposed recognition tool showed to be cost-effective, and highly accurate for the real-time gait analysis at different walking conditions.

This paper deals with optimal robot assistance based on estimation of torque and impedance parameters of patients performing robot-aided rehabilitation of ankle movements. A generalized momenta-based disturbance observer is used to estimate the patient’s torque. From this estimate, the stiffness and damping parameters are determined by the least squares method, considering the patient motor control is modeled as an impedance control. An optimal control strategy is proposed to properly define the robot assistance during therapy session. Experimental results considering a healthy subject wearing an ankle robot and performing a set of movements are presented.

Inertial Measurements Unit (IMU) based systems are a purposeful and alternative tool to monitor human gait mainly because they are cheaper, smaller and can be used without space restrictions compared to other gait analysis methods. In the scientific community, there are well-known studies that test the accuracy and efficiency of this method compared to ground truth systems. Gait parameters such as stride length, distance, velocity, cadence, gait phases duration and detection, or joint angles are tested and validated in these studies in order to study and improve this technology. In this article, knee joint angles were calculated from IMUs’ data and they were compared with DARwIn OP knee joint angles. IMUs were attached to the left leg of the robot and left knee flexion-extension (F-E) was evaluated. The RMSE values were less than 6° when DARwIn OP was walking, and less than 5° when the robot kept the left leg stretched and performed an angle of -30°.

The development of robust algorithms for human gait analysis are essential to evaluate the gait performance, and in many cases, crucial for diagnosing gait pathologies. This work proposes a new adaptive tool for human gait event detection in real-time, based on the angular velocity recorded from one gyroscope placed on the instep of the foot and in a finite state machine with adaptive decision rules. The signal was segmented to detect 6 events: Heel Strike (HS), Foot Flat (FF), Middle Mid-Stance (MMST), Heel-Off (HO), Toe-Off (TO), and Middle Mid-Swing (MMSW). The tool was validated with healthy subjects in ground-level walking using a treadmill, for different speeds (1.5 to 4.5 km/h) and slopes (0 to 10%). The results show that the tool is highly accurate and versatile for the detection of all events, as indicated by the values of accuracy, average delays and advances (HS: 99.96%, -7.95 ms, and 9.85 ms; FF: 99.48%, -4.95 ms, and 9.35 ms; MMST: 98.26%, - 36.54 ms, and 16.38 ms; HO: 98.87%, -22.71 ms, and 18.62 ms; TO: 95.95%, -6.80 ms, 14.38 ms; MMSW: 96.06%, -3.45 ms; 0.15 ms, respectively). These findings suggest that the proposed tool is suitable for the real-time gait analysis in real-life activities.

Cerebral Palsy (CP) is the most common cause of permanent serious physical disability in childhood. CPWalker is a robotic platform through which the child can start experiencing autonomous locomotion in a rehabilitation environment. The main objective of this work is to biomechanically compare difference conditions of gait assistance in Children with CP using CPWalker. The results showed some differences among three patients with different conditions in terms of gait parameters, pelvic angles and propulsion. The observed changes during the therapy show the potential of this rehabilitation device.

Parkinson’s Disease is a neurodegenerative disorder for which there is still no cure affecting the non-motor and motor systems. One of the most serious gait disorders are the freezing episodes, denominated by freezing of gait. This paper address the development and validation of a neurofeedback vibrotactile system through a belt for patients with Parkinson’s Disease overcome the freezing episodes, aiming to detect the most perceived frequency. It was verified that the higher frequencies (above 160 Hz) are easily perceived independently of the time interval of vibrotactile feedback.

There are more than one million people, each year, that will suffer a lower limb amputation. This condition occurs as a result of a wide range of diseases: diabetes, trauma or malignant tumors. An amputation means disability and a poor quality of life. The major challenge in the development of a prosthesis lies in restoring the missing human function, i.e. locomotion while maintaining the biomechanical requirements of the ankle joint. Thus, this work addresses the field of the artificial devices that replace the ankle-foot. The goal of this work is to sketch a 3D model of a transtibial prosthesis, and to find and implement a dynamic model of human ankle joint motion, in order to be used in a future control system strategy. Thereby, this work represents the first steps towards the development of a transtibial prosthesis.

Each year thousands of people suffer from lower limb amputation, mainly due to three causes: wars, accidents and vascular diseases. The development of lower limb prosthesis is crucial to restore people’s mobility, improving the quality of life of millions of people. This contribution presents a gait events detector algorithm capable of detecting all the events of the human walking. Results show the correct transition between phases during several gait cycles for a human model walking in flat terrain.

The following section is included:

• AUTHOR INDEX