We aim to reliably predict whether a grasp on a known object is successful before it is executed in the real world. There is an entire suite of grasp metrics that has already been developed which rely on precisely known contact points between object and hand. However, it remains unclear whether and how they may be combined into a general purpose grasp stability predictor. In this paper, we analyze these questions by leveraging a large scale database of simulated grasps on a wide variety of objects. For each grasp, we compute the value of seven metrics. Each grasp is annotated by human subjects with ground truth stability labels. Given this data set, we train several classification methods to find out whether there is some underlying, non-trivial structure in the data that is difficult to model manually but can be learned. Quantitative and qualitative results show the complexity of the prediction problem. We found that a good prediction performance critically depends on using a combination of metrics as input features. Furthermore, non-parametric and non-linear classifiers best capture the structure in the data.

Many analytical metrics have been proposed to evaluate the quality of a grasp based on different criteria and principles. To use most of them in practical real applications, some operational parameters need to be determined: maximum and minimum values, normalization ratios, quality thresholds, robustness in front of position errors and, more importantly, relations between alternative metrics. This paper proposes a methodology to study and characterize the operational parameters that allow the use of several metrics in practical applications, and comparing them. The proposed approach uses exhaustive simulation testing to obtains statically significant results regarding the measurements of several quality metrics. This allows an informed setting of the practical operational values for each metric. Results are provided for a Barrett hand grasping a varied set of objects.

Prosthetic hands have evolved and improved over the years, helping people gaining manipulation capabilities. Having a simulation tool able to obtain quantitative evaluation of the grasp capabilities of such hands could give insights as how to improve the design of hand prostheses or robotic hands by means of obtaining better quality scores. The purpose of this work is to present a framework developed to evaluate the grasp capabilities of a prosthetic hand using a selected set of grasp quality measures, and compare the results with the ones obtained for the human hand using a biomechanical model. Experiments grasping an object with different postures and varying aspects of the prosthetic hand model were performed showing the functionality of the proposed framework to evaluate the grasp quality.