Robot Experiments Research Articles

An adaptive bilateral impedance control based on nonlinear disturbance observer for different flexible targets grasping

The uncertainty of environmental parameters and time-varied characteristics usually generate contact force errors resulting the inaccuracy or even task failed in many tele-operation systems. In order to solve the problem, a novel composite bilateral tele-operation controller is proposed in this paper combining adaptive impedance control with nonlinear disturbance observer-based sliding mode controller. An adaptive impedance controller is constructed to achieve high-precision tracking of desired contact forces and ensure the safety of grasping various flexible targets. A sliding mode controller on the basis of a nonlinear disturbance observer is designed, so as to accurately estimate the composite disturbance signals as well as to ensure the stability and the high accuracy of the trajectory tracking. The stability of the system is proved by the Lyapunov function. Numerical simulations and real robot experiments are performed. Results shows that our method can greatly enhance the safety of the flexible target in the process, minimize the influence of disturbance signals on the system, and improve the robustness of the system.

Autonomous optimization of non-aqueous Li-ion battery electrolytes via robotic experimentation and machine learning coupling

Developing high-energy and efficient battery technologies is a crucial aspect of advancing the electrification of transportation and aviation. However, battery innovations can take years to deliver. In the case of non-aqueous battery electrolyte solutions, the many design variables in selecting multiple solvents, salts and their relative ratios make electrolyte optimization time-consuming and laborious. To overcome these issues, we propose in this work an experimental design that couples robotics (a custom-built automated experiment named "Clio”) to machine-learning (a Bayesian optimization-based experiment planner named "Dragonfly”). An autonomous optimization of the electrolyte conductivity over a single-salt and ternary solvent design space identifies six fast-charging non-aqueous electrolyte solutions in two work-days and forty-two experiments. This result represents a six-fold time acceleration compared to a random search performed by the same automated experiment. To validate the practical use of these electrolytes, we tested them in a 220 mAh graphite∣∣LiNi0.5Mn0.3Co0.2O2 pouch cell configuration. All the pouch cells containing the robot-developed electrolytes demonstrate improved fast-charging capability against a baseline experiment that uses a non-aqueous electrolyte solution selected a priori from the design space.

A reinforcement learning fuzzy system for continuous control in robotic odor plume tracking

AbstractIn dynamic outdoor environments characterized by turbulent airflow and intermittent odor plumes, robotic odor plume tracking remains challenging, because existing algorithms heavily rely on manually tuning or learning from expert experience, which are hard to implement in an unknown environment. In this paper, a multi-continuous-output Takagi–Sugeno–Kang fuzzy system was designed and tuned with reinforcement learning to solve the robotic odor source localization problem in dynamic odor plumes. Based on the Lévy Taxis plume tracking controller, the proposed fuzzy system determined the parameters of the controller based on the robot’s observation and guided the robot to turn and move towards the odor source at each searching step. The trained fuzzy system was tested in simulated filament-based odor plumes dispersed by a changing wind field. The results showed that the performance of the proposed fuzzy system-based controller trained with reinforcement learning can achieve a similar success rate and higher efficiency compared with a manually tuned and well-designed fuzzy system-based controller. The fuzzy system-based plume tracking controller was also validated through real robotic experiments.

A flexible wearable e-skin sensing system for robotic teleoperation

AbstractElectronic skin (e-skin) is playing an increasingly important role in health detection, robotic teleoperation, and human-machine interaction, but most e-skins currently lack the integration of on-site signal acquisition and transmission modules. In this paper, we develop a novel flexible wearable e-skin sensing system with 11 sensing channels for robotic teleoperation. The designed sensing system is mainly composed of three components: e-skin sensor, customized flexible printed circuit (FPC), and human-machine interface. The e-skin sensor has 10 stretchable resistors distributed at the proximal and metacarpal joints of each finger respectively and 1 stretchable resistor distributed at the purlicue. The e-skin sensor can be attached to the opisthenar, and thanks to its stretchability, the sensor can detect the bent angle of the finger. The customized FPC, with WiFi module, wirelessly transmits the signal to the terminal device with human-machine interface, and we design a graphical user interface based on the Qt framework for real-time signal acquisition, storage, and display. Based on this developed e-skin system and self-developed robotic multi-fingered hand, we conduct gesture recognition and robotic multi-fingered teleoperation experiments using deep learning techniques and obtain a recognition accuracy of 91.22%. The results demonstrate that the developed e-skin sensing system has great potential in human-machine interaction.

Semi-empirical dynamics modeling of a bicycle robot based on feature selection and RHONN

Aiming at the difference between a bicycle robot dynamics model based on prior knowledge and a real robot, the robot modeling problem is studied. A set of features is extracted from the dynamics model of a bicycle robot. The trigonometric functions in the features are replaced with a linear combination of several sigmoid functions to form a new set of features. An index describing the importance of features is designed based on the maximal information coefficient (MIC). The features are ranked using the feature importance index. The index of system identification fitness (FIT) is regarded as the likelihood function term in Bayesian information criterion (BIC). The penalty term of BIC is constructed with the number of both the features and weight parameters in the recurrent high order neural network (RHONN) model. A semi-empirical dynamics modeling method based on feature selection is proposed. An inverse optimal controller is designed to stabilize the bicycle robot. The effectiveness of the modeling method is verified through simulation and real robot experiments.

Dances with Social Robots: A Pilot Study at Long-Term Care

Dance therapy can have significant physical, emotional and cognitive benefits for older adults. In particular, social robots can be developed to autonomously facilitate dance sessions to engage these individuals with the aim of improving quality of life. To successfully integrate and promote long-term use of social robots into long-term care homes for such recreational activities, it is important to explore both residents’ and staff’s perceptions of such robots. In this paper, we present the first pilot human–robot interaction study that investigates the overall experiences and attitudes of both residents and staff in a long-term care home for robot-facilitated dance sessions. In general, the questionnaire results from our study showed that both staff and residents had positive attitudes towards the robot-facilitated dance activity. Encouraging trends showed residents had higher ratings for statements on perceived ease of use, safety, and enjoyment than the staff. However, the staff had a statistically significantly higher rating for willingness to use the robots for dance facilitation. Some key statistical differences were also determined with respect to: (1) gender within the resident group (men had higher ratings for the robots being useful in helping facilitate recreational activities), as well as between staff and residents (resident men had higher perceived safety), and (2) prior robot experience (residents with limited prior experience had higher ratings on perceived ease of use and perceived enjoyment than staff with the same level of experience). The robot-facilitated dance activity was positively received by both older adults and staff as an activity of daily living that can enhance wellbeing while also being safe, easy to use and enjoyable.

Design of a Variable Stiffness Gecko-Inspired Foot and Adhesion Performance Test on Flexible Surface.

Adhesion robots have broad application prospects in the field of spacecraft inspection, repair, and maintenance, but the stable adhesion and climbing on the flexible surface covering the spacecraft has not been achieved. The flexible surface is easily deformed when subjected to external force, which makes it difficult to ensure a sufficient contact area and then detach from it. To achieve stable attachment and easy detachment on the flexible surface under microgravity, an adhesion model is established based on the applied adhesive material, and the relationship between peeling force and the rigidity of the base material, peeling angle, and working surface stiffness is obtained. Combined with the characteristics of variable stiffness structure, the adhesion and detachment force of the foot is asymmetric. Inspired by the adhesion-detachment mechanism of the foot of the gecko, an active adhesion-detachment control compliant mechanism is designed to achieve the stable attachment and safe detachment of the foot on the flexible surface and to adapt to surfaces with different rigidity. The experimental results indicate that a maximum normal adhesion force of 7.66 N can be generated when fully extended, and the safe detachment is achieved without external force on a flexible surface. Finally, an air floating platform is used to build a microgravity environment, and the crawling experiment of a gecko-inspired robot on a flexible surface under microgravity is completed. The experimental results show that the gecko-inspired foot with variable stiffness can satisfy the requirements of stable crawling on flexible surfaces.

Cognitive load affects early processes involved in mentalizing robot behaviour

How individuals interpret robots’ actions is a timely question in the context of the general approach to increase robot’s presence in human social environment in the decades to come. Facing robots, people might have a tendency to explain their actions in mentalistic terms, granting them intentions. However, how default or controllable this process is still under debate. In four experiments, we asked participants to choose between mentalistic (intentional) and mechanistic (non-intentional) descriptions to describe depicted actions of a robot in various scenarios. Our results show the primacy of mentalistic descriptions that are processed faster than mechanistic ones (experiment 1). This effect was even stronger under high vs low cognitive load when people had to decide between the two alternatives (experiment 2). Interestingly, while there was no effect of cognitive load at the later stages of the processing arguing for controllability (experiment 3), imposing cognitive load on participants at an early stage of observation resulted in a faster attribution of mentalistic properties to the robot (experiment 4). We discuss these results in the context of the idea that social cognition is a default system.

Safe Technical Innovation: Development and Implementation of a Robotic Breast Operation Program.

Safe Technical Innovation: Development and Implementation of a Robotic Breast Operation Program.

Design, Analysis and Experiments of Hexapod Robot with Six-Link Legs for High Dynamic Locomotion.

An important feature of a legged robot is its dynamic motion performance. Traditional methods often improve the dynamic motion performance by reducing the moment of inertia of robot legs or by adopting quasi-direct drive actuators. This paper proposes a method to enhance the dynamic performance of a legged robot by transmission mechanism. Specifically, we present a unique six-link leg mechanism that can implement a large output motion using a small drive motion. This unique feature can enhance the robots’ dynamic motion capability. Experiments with a hexapod robot verified the effectiveness of the mechanism. The experimental results showed that, when the steering gear of the robot rotates 1°, the toe can lift 7 mm (5% of body height), and the maximum running speed of the robot can reach 390 mm/s (130% of the moveable body length per second).

Robotic simulation: validation and qualitative assessment of a general surgery resident training curriculum.

The da Vinci skills simulation curriculum has been validated in the literature. The updated simulator, SimNow, features restructured exercises that have not been formally validated. The purpose of this study is to validate the SimNow resident robotic basic simulation curriculum. This study also consists of a qualitative assessment that gives greater insight into the learner's experience completing the robotic curriculum. There were 18 participants in this study: 6 novices, 6 competent surgeons, and 6 expert surgeons. The curriculum comprised 5 exercises; participants completed three consecutive scored trials. Computer-derived performance metrics were recorded. The NASA Task Load Index survey was used to assess subjective mental workload. Subjects were asked a series of open-ended questions regarding their experience that were recorded and transcribed. Codes were identified using an inductive method, and themes were generated. Performance metrics were significantly different between novice versus competent and expert surgeons. There was no significant difference in any score metric between competent and expert surgeons. On average, overall score percentages for competent and expert surgeons were between 90.4 and 92.8% versus 70.5% for novices (p = 0.02 and p = 0.01). Expert surgeons perceived a higher level of performance completing the exercises than novice surgeons (15.8 vs. 45.8, p = 0.02). Participants noted a similar robotic experience, utilizing efficiency of motion and visual field skills. Participants agreed on exercise strengths, exercise weaknesses, and software limitations. Competent and expert surgeons were better able to assess the exercises' clinical application. The SimNow curriculum is a valid simulation training as part of a general surgery resident robotic curriculum. The curriculum distinguishes between novices compared to competent and expert surgeons, but not between competent and expert surgeons. Clinical training level does not affect the experience and mental workload using the robotic simulator, except for competent and expert surgeons' ability to better assess clinical application.

Redundant Robot with Pneumatic Artificial Muscles for Rehabilitation Works Using Iterative Learning Control

This study aims to develop a multi-functional robotic arm with a multi-degree of freedom using a pneumatic artificial muscle cylinder as the main actuator, with a wearable function. While wearing the robotic arm, it has four degrees of freedom and can be used as an upper limb rehabilitation aid to perform rehabilitation exercise tasks. The rotating axes are driven by pneumatic artificial muscle cylinders to ensure the safety and flexibility of the robot interaction. By integrating multiple pneumatic control valves and sensors with embedded interface cards, the control of pneumatic artificial muscle cylinders is performed, and the rehabilitation trajectory commands are planned through the kinematics of the robotic arm, and a closed-loop control system is established to enable the robotic arm to achieve the task of rehabilitation trajectory tracking. In the single-axis robotic arm controller experiment, it was found that the LMS-PID is superior to the conventional PID control method. In the control experiment of the robotic arm under the multi-axis recovery trajectory, the result showed that the training of drawing a circle on the wall could be successfully planned. In the multi-axis robotic tracking experiment under the rehabilitation trajectory, the result showed that the RMSE of the tracking trajectory decreased to 0.2444 and 0.2853. In the robotic arm joining/loading experiment, it was shown that the method of this study can withstand some loading effects. Moreover, this study introduces iterative learning control to improve the non-linear compensation and phase lag problems of the PID controller so that the robot arm can have a certain tracking accuracy under the round-trip robust trajectory.

Effects of the Mechanical Closed-Loop Between the Body and the Ground on the Postural Balance of Gaits

People and animals adapt their gait to the environment as they perform activities in a variety of environments. However, there are cases where the parts of the body necessary for walking are damaged in some way, resulting in walking difficulties. An example is paralysis caused by a stroke. A split-belt treadmill is occasionally used for the investigation to analyze how the stroke effects on the motion. However, the mechanical properties of the split-belt treadmill on the body have not been clarified. It is also unknown how the mechanical closed-loop between the body and the environment, generated by synchronizing the movements of the two belts, affects the gait. In this study, we investigated that the effect of the mechanical closed-loop structure between the body and the environment on walking using the robot and the mechanical effect of the floor reaction force on the body. Further, we conducted walking experiments using the developed robot, obtained body and environmental information, and analyzed the results. As the result, it was observed that the motion data differed based on the coupling of the treadmill. In other words, it was suggested that the mechanical closed-loop structure certainly influenced the physical balances on walking motion. Furthermore, it is confirmed that the coupling of treadmills increases the body’s sway. Although our results are given from a robotic experiment, it is expected that these measures would be one of the important index in human rehabilitations.

Walking experiments of small and lightweight rimless wheel robot

Walking experiments of small and lightweight rimless wheel robot

Factor graph-based high-precision visual positioning for agricultural robots with fiducial markers

High-precision positioning of agricultural robots is the key to the automation of greenhouse agricultural production. We propose a visual positioning method by leveraging the fiducial markers and factor graph to achieve reliable performance. This method consists of two parts, a front-end module and a back-end module. First, fiducial markers are involved to the front-end module to tackle the problems of unstructured and insufficient features in the greenhouse environment. A tag detection algorithm AprilTag2 is used to identify markers and extract feature points for the positioning purpose. Second, to eliminate the adverse effects of camera motion, mechanical vibration, and environmental disturbance on the positioning accuracy, the back-end module considers the characteristics of robot motion and the constraint relationship of variables, and correspondingly designs a positioning model based on the factor graph. The model combines five factors for maximum a posteriori estimation and incremental optimization of the state of the robot in real time. Finally, the positioning experiment of the wheeled robot was carried out at the motion speed of 0.26 m/s, 0.39 m/s, and 0.52 m/s. The experimental results show that the average errors of positioning are 0.056 m, 0.065 m, and 0.081 m, respectively, and the standard deviations of the errors are lower than 0.05 m. The visual positioning method based on fiducial markers and factor graph outperforms the mainstream positioning techniques in terms of positioning accuracy and robustness, and meet the high-precision positioning requirements in complex agricultural fields.

Deep reinforcement learning in playing Tetris with robotic arm experiment

Tetris has been an important field for research in deep reinforcement learning (DRL). However, most studies about Tetris are focused on simulation validation, and a few attempts are conducted in the real-world environment. In this paper, the DRL algorithms are trained in the constructed Tetris simulation environment, after that they are deployed into the real-world Tetris experiments. The dynamic timesteps method is integrated into the proximal policy optimization (PPO) method to accelerate its training speed, which reaches the goal of the game within 1483 episodes. With the help of multiple recognition and segmented moving techniques, the robotic arm provides accurate and robust performance to play real-world Tetris. The effectiveness of the developed system is experimentally verified; the experimental results show that the proposed algorithm achieved superior performance compared with conventional method and Deep Q-Network (DQN) in real-world Tetris environments.

Effect of variable-angle trajectory structure on mechanical performance of CF/PEEK laminates made by robotic fiber placement

In comparison to the traditional fixed-angle trajectory, the variable-angle trajectory has a greater design space. However, it is a challenge to determine which common design curve structure is the most effective for improving mechanical performance. This work explores the effects of various design curves such as fixed-angle curve, linear curve, arc curve, sine curve, Bezier curve, and cubic polynomial curve trajectories on mechanical performance of laminates, including vibration modal performance and buckling performance. Genetic algorithm and improved NSGA-II algorithm are then used to optimize various curve structures. The results are confirmed utilizing thermoplastic Carbon Fiber (CF)/Polyether-Ether-Ketone (PEEK) laminates made by robotic fiber placement experiments. The relationship model between different mechanical performance and curve design variables is established. The optimization of variable-angle structure with mechanical properties as input variables is achieved. Meanwhile, a full-process angle-variable laying software platform from trajectory planning (CAD), trajectory optimization to manufacturing (CAM) is developed for facilitating the fiber placement application.

The Ability of Robot-Assisted Total Knee Arthroplasty in Matching the Efficiency of Its Conventional Counterpart at an Orthopaedic Specialty Hospital

BackgroundRobot-assisted total knee arthroplasty (RA-TKA) has become a popular technology. Studies have investigated the learning curve for surgeons incorporating RA-TKA into practice, but less is known regarding the change in operative efficiency when introducing RA-TKA into a facility. The purpose of this study was to investigate the effects of RA-TKA on operative and turnover time at an orthopaedic specialty hospital. MethodsA total of 148 cases (74 RA-TKA and 74 conventional TKA [C-TKA]) performed by 2 surgeons with previous robotic experience were identified following the introduction of RA-TKA at our facility. Patient demographics, comorbidities, and operative times (ie, wheels-in to incision, incision to closure, closure to wheels-out, and turnover time) were recorded. Cumulative summation analyses were used to investigate learning curves of factors extraneous to surgeon proficiency with RA-TKA. ResultsWhile RA-TKA had a slightly longer set up (3 minutes; range, 12-45), surgical (5 minutes; range, 33-118), and breakdown time (3 minutes; range, 2-7), there was no difference in turnover time between the groups. The learning curve for surgeon A was 6 robotic cases, whereas surgeon B demonstrated no learning curve. There was no identifiable learning curve for turnover time. ConclusionThere was a mean of 8 minutes of increased time required to perform a RA-TKA compared to C-TKA. However, these small increases for the RA-TKA group for set-up, surgical, and breakdown times are not likely to be clinically relevant compared to the C-TKA. It appears that the RA-TKA technology was able to be incorporated into this specialty hospital with minimal changes to surgical efficiency.

Mirror-Based Robotic Therapy for Ankle Recovery with a Serious Game: A Case Study with a Neurological Patient.

Neuromuscular disorders, such as foot drop, severely affect the locomotor function and walking independence after a brain injury event. Mirror-based robotic therapy (MRT) has been a promising rehabilitation strategy favouring upper limb muscle strength and motor control in the last years. However, there are still no studies validating this technique in lower limb experimental protocols. This paper presents an innovative visual and motor feedback strategy based on serious games and MRT modalities. Thus, a preliminary system validation with a healthy participant is performed. Moreover, the strategy's potential effects were investigated in a neurologic patient's short rehabilitation program. After six sessions, the results of the method favoured active ankle plantarflexion range of motion and muscle activation. Although the patient had a positive adaptation at the end of the game, it is necessary to improve the proposed strategy to enhance the robotic experience in the long term.

A comparison of robotic versus laparoscopic distal pancreatectomy: a single surgeon's robotic experience in a high-volume center.

Robotic surgery is the most recent advanced minimally invasive approach for distal pancreatectomy. However, its benefits over laparoscopic distal pancreatectomy (LDP) remain undetermined. Previous studies were limited by their small sample size or variations in surgeon skills. This study aimed to compare robotic distal pancreatectomy (RDP) performed by a single surgeon with LDP performed by skilled laparoscopic surgeons in a high-volume center. We retrospectively analyzed consecutive RDP performed by a single surgeon between December 2020 and November 2021 with LDP performed by experienced surgeons during the same period in a high-volume center. Patient characteristics and perioperative variables were compared. The analysis included 55 RDP and 146 LDP procedures. The operative time in the RDP group was significantly shorter than the LDP group (171 vs. 222min, P < 0.001), both in spleen-preserved (154 vs. 212min, P < 0.001) and spleen-removed (192 vs. 230min, P = 0.005) procedures. The RDP group made more frequent use of the stapler technique for pancreas transection (87.3 vs. 68.5%, P = 0.007), and its estimated blood loss was lower (79 vs. 155mL, P < 0.001) than the LDP group. The postoperative hospital stay in the RDP group was significantly shorter than the LDP group (8 vs. 12days, P < 0.001). The groups were similar in their complication distributions. RDP is as safe and feasible a minimally invasive approach as LDP. The advanced manipulation and visualization capabilities of the robotic approach in distal pancreatectomy could help reduce operative time and blood loss, and is related to shorter postoperative hospital stay.