50 research outputs found
Experimental Evaluation of Human Grasps Using a Sensorized Object
No full textRoa M, Kõiva R, Castellini C. Experimental Evaluation of Human Grasps Using a Sensorized Object. Presented at the IEEE International Conference on Biomedical Robotics and Biomechatronics (BioRob 2012), Rome, Italy.Grasp quality measures have been studied for long time, given their importance to evaluate the goodness/convenience of a grasp made with a robotic hand. However, the application of these quality measures to the grasps made by humans has just recently received some attention. This paper presents an experimental evaluation and comparison of different measures, using data obtained with a sensorized object. The experiment compares power grasps and precision grasps obtained with different number of fingers. The results intend to be a guide to the application of such qualities in the evaluation of robotic grasp actions
Passive safe falling of humanoid robots
No full textLAUREA MAGISTRALEIn questa tesi è riportato lo studio del problema della ‘Caduta Passiva Sicura’ del robot umanoide TORO (TOrque controlled humanoid RObot) sviluppato al DLR (Centro Spaziale Tedesco) di Oberpfafenhoffen, in Germania.
Il termine ‘Caduta Passiva’ si riferisce a tutti quei casi dove il controllo del robot durante una caduta risulta difficile se non addirittura impossibile. Dato che non è possibile controllare la caduta, l’unica modalità rimasta per renderla ‘sicura’ risulta essere l’implementazione di protezioni. Dopo aver analizzato a fondo le caratteristiche di TORO e aver studiato lo stato dell’arte riguardante questo problema, risulta che gli airbag siano la tecnologia che meglio sposa i prerequisiti di questo tipo di applicazione.
L’idea è quella di realizzare una campagna sperimentale per studiare sia l’entità del fenomeno in sé (massime forze e accelerazioni) sia l’efficacia dell’utilizzo di airbag. Poiché risulta essere troppo pericoloso operare i test direttamente su TORO, non essendo esso progettato per sostenere cadute, viene progettata e costruita una copia cinematica e inerziale del robot in scala 1:1. Successivamente, vengono realizzate e implementate diverse soluzioni airbag. L’apparato è poi testato facendolo cadere in pose differenti con e senza airbag e le storie temporali di accelerazione del bacino e del petto vengono misurate attraverso accelerometri.
Unendo la conoscenza acquisita attraverso i test con relazioni teoriche, viene sviluppato un modello analitico del fenomeno. Il modello crea una rappresentazione 3D del robot, che replica la distribuzione del volume e della massa. Queste informazioni vengono utilizzate per calcolare la direzione, i possibili punti di contatto e la dinamica della caduta. Usando la stima della dinamica, è possibile calcolare durata, decelerazioni e forze coinvolte nel fenomeno.
I risultati del modello vengono poi confrontati con i risultati sperimentali. Questo tipo di modello può essere utilizzato come struttura di base sia per il progetto di airbag migliori sia per l’implementazione di strategie di controllo di gonfiaggio/sgonfiaggio degli airbag sia infine per stimare forze e decelerazioni per casi differenti da quelli testati sperimentalmente. Il modello può essere inoltre utilizzato come punto di partenza per progettare in futuro robot più resistenti.This thesis reports the study of the problem of passive safe falling of the humanoid robot TORO (TOrque controlled humanoid RObot) developed by DLR (German Aerospace Center) in Oberpfafenhoffen, Germany.
Passive falling refers to all those cases when control of the robot during fall is difficult or even not possible. Since the control is not available, this kind of phenomenon can be made ‘safe’ only through the implementation of protective gear. Analyzing deeply the characteristics of TORO and studying the state of the art of the problem, airbags are individuated as the technology that best matches the requirements of this kind of application.
The idea is to develop an experimental campaign to study the phenomenon and the effectiveness of airbags. Since it is too dangerous to make the experiments directly with TORO, as it was not designed for safe falling, a mockup of the robot is designed and built to match both kinematic and inertial characteristics. Then different airbag solutions are realized and implemented. The mockup is then tested by falling in different poses with and without airbags and accelerations of hip and chest are acquired through accelerometers.
Merging the knowledge acquired from experiments with theoretical relations we develop an analytical model to describe the phenomenon. The model is a 3D representation of the robot, replicating its volume and mass distribution. This information is then used to identify direction of fall, contact points and dynamics of the fall and impacts. Using the estimated dynamics, we can calculate the times, decelerations and forces involved in the fall.
The model results are then compared with the experimental ones. This kind of model can be used as a framework for improvement of airbag design, inflation/deflation control logics and estimation of forces/decelerations for cases different than the tested ones. The learnings can potentially be used as a basis to design future more robust humanoids
Emerging Paradigms for Robotic Manipulation: from the Lab to the Productive World
No full textThe articles in this special section aim to stimulate and gather publications describing how new approaches in the field of robotic manipulation can be (or have already been) transferred from research labs to the productive world. Novel robotic developments are going hand in hand with the need for innovation in manufacturing and service applications. Large industries increasingly need fast-adapting production lines, whereas small and medium-sized enterprises are adopting collaborative manipulators to gain a competitive edge in global markets. Additionally, a growing number of companies are producing and using service robots for a variety of applications, from agriculture to surgery. In this emerging framework, robots should be able to grasp and manipulate different tools as well as operate in fast-changing or even unstructured surroundings, possibly interacting with human users or coworkers. These real-world scenarios require going beyond traditional parallel-jaw grippers and preprogrammed or teleoperated trajectories and require searching for innovative grasping and manipulation paradigms that can cope with the needs of uncertain environments
Stable Myoelectric Control of a Hand Prosthesis using Non-Linear Incremental Learning
No full textStable myoelectric control of hand prostheses remains an open problem. The only successful human-machine interface is surface electromyography, typically allowing control of a few degrees of freedom. Machine learning techniques may have the potential to remove these limitations, but their performance is thus far inadequate: myoelectric signals change over time under the influence of various factors, deteriorating control performance. It is therefore necessary, in the standard approach, to regularly retrain a new model from scratch.We hereby propose a non-linear incremental learning method in which occasional updates with a modest amount of novel training data allow continual adaptation to the changes in the signals. In particular, Incremental Ridge Regression and an approximation of the Gaussian Kernel known as Random Fourier Features are combined to predict finger forces from myoelectric signals, both finger-by-finger and grouped in grasping patterns.We show that the approach is effective and practically applicable to this problem by first analyzing its performance while predicting single-finger forces. Surface electromyography and finger forces were collected from 10 intact subjects during four sessions spread over two different days; the results of the analysis show that small incremental updates are indeed effective to maintain a stable level of performance.Subsequently, we employed the same method on-line to teleoperate a humanoid robotic arm equipped with a state-of-the-art commercial prosthetic hand. The subject could reliably grasp, carry and release everyday-life objects, enforcing stable grasping irrespective of the signal changes, hand/arm movements and wrist pronation and supination
Simple passive dynamic models for emerging quadrupedal gaits
No full textLAUREA MAGISTRALEOggigiorno i robot dotati di zampe sono preferiti ai rover per applicazioni come mappatura ed esplorazione di ambienti non strutturati. Solitamente, questi sistemi eseguono i movimenti richiesti grazie ad un controllo appropriato. Tuttavia, recenti sviluppi suggeriscono che soluzioni ad alta efficienza energetica possono essere ottenute da sistemi con elasticità che esibiscono naturalmente specifici pattern motori. In ciò è stata presa ispirazione dagli animali, i quali sfruttano vari pattern motori per muoversi a diverse velocità minimizzando il dispendio energetico. Modelli dinamici semplificati sono risultati utili per la comprensione della locomozione in natura. Il più comune, il modello SLIP (Spring Loaded Inverted Pedulum) riesce a spiegare gli effetti dell'elasticità in zampe lineari. Sfortunatamente, questo modello risulta inappropriato per lo studio della stabilità dei cicli limite in sistemi con arti segmentati. L'obiettivo di questo lavoro è rivisitare modelli dinamici semplificati per robot quadrupedi con elasticità e caratterizzare la loro stabilità per sfruttare piu facilmente la loro dinamica naturale al fine di favorire la locomozione. Vengono introdotti due semplici modelli dinamici con gambe elastiche segmentate. Un primo modello sfrutta l’elasticità solo nella direzione radiale della gamba, mentre il secondo anche in quella torsionale. Il loro comportamento dinamico è analizzato attraverso simulazioni, con un approccio di continuazione numerica delle soluzioni. Dalla dinamica passiva di questi sistemi, i pattern motori emergono senza imporre sequenze di contatto a priori. Conducendo un'analisi parametrica dei più importanti pattern motori trovati, è stato identificato l'effetto dei principali parametri di sistema sulla stabilità dei cicli limite. I risultati ottenuti possono aiutare nella progettazione di robot quadrupedi che sfruttano naturalmente l'intrinseca elasticità di sistema.Nowadays legged robots are being preferred to wheeled vehicles for applications like mapping and exploration of unstructured environments. Usually, these mechanical systems perform the required motions thanks to suitable controllers. However, recent works suggest that energy-efficient solutions can be obtained by designing elastic systems that naturally exhibit locomotion patters. Inspiration is taken from nature and biological legged locomotion. Animals use different gaits to move at certain speeds while minimizing energy consumption. To understand the underlying dynamics of biological locomotion, simplified models have been proposed. The most common one, the SLIP (Spring Loaded Inverted Pendulum) model, can explain the effect of the radial elasticity of linear legs and generate patterns for locomotion of different legged systems. Unfortunately, it is inappropriate for the study of stability of limit cycles in systems with articulated legs. The aim of this work is to revisit simplified dynamic models for quadrupedal elastic robots, and characterize their stability in order to facilitate the exploitation of their natural dynamics for locomotion purposes. Two simple passive quadrupedal models featuring segmented elastic legs are introduced. One model exploits only leg radial elasticity, while the other model exploits also the rotational one. The dynamic behavior of these models is analyzed using numerical simulation and a continuation approach. From the passive dynamics of such models, gaits emerge without imposing contact sequences a priori. Conducting a parametric analysis on the most important gaits found on this model, the effect of the main system design parameters on gait stability has been identified. The obtained results help in the design of elastic quadrupedal robots that naturally exploit inherent system elasticity
Respiratory Muscles: Structure, Function and Relationship with the ACE Gene. A Brief Morphofunctional Communication
Get PDFGrasp Quality Evaluation Done Right: How Assumed Contact Force Bounds Affect Wrench-Based Quality Metrics
Get PDFWrench-based quality metrics play an important role in many applications such as grasp planning or grasp success prediction. In this work, we study the following discrepancy
which is frequently overlooked in practice: the quality metrics are commonly computed under the assumption of
sum-magnitude bounded contact forces, but the corresponding grasps are executed by a fully actuated device where the contact forces are limited independently. By means of experiments carried out in simulation and on real hardware, we show that in this setting the values of these metrics are severely underestimated. This can lead to erroneous conclusions regarding the actual capabilities of the grasps under consideration. Our findings highlight the importance of matching the physical properties of the task and the grasping device with the chosen quality metrics
Quadrupedal template model for parametric stability analysis of trotting gaits
No full textSimple template models have proven useful for understanding the underlying dynamics of legged locomotion. The most common one, the SLIP model, considers the legs as linear springs with constant stiffness, and it explains well the radial dynamics of the legs. However, in order to study the influence of the leg swing dynamics and leg segmentation on gait stability, more complex models are required. This paper introduces a novel template model for quadrupedal gait, which considers these additional aspects. The dynamic behavior of the model is analyzed via numerical simulation, using a continuation approach. By conducting a parametric analysis on the trotting gait and analyzing its stability, we identify the influence of the main model parameters, leading to marginally unstable limit cycles. These numerical results are applicable to the design of more efficient elastic quadrupedal robots
Automated Planning of Workcell Layouts Considering Task Sequences
No full textThe initial design of a robotic workcell layout has a large impact on the feasibility and performance of the intended robotic tasks. We define this layout design as a constrained nonlinear optimization problem that aims to optimize the placement of workcell components by minimizing the distance traveled between task sequences while maximizing the robot's manipulability. Suitable constraints guarantee the reachability as well as the absence of collisions. We solve this optimization problem via a genetic algorithm, and demonstrate it in three scenarios for a dual-arm robotic system that assembles product variants out of aluminum profiles
