The Hierarchical Agent Control Architecture (HAC) is a general toolkit for specifying an agent's behavior. By organizing the hierarchy around tasks to be accomplished, not the agents themselves, it is easy to incorporate multi-agent actions and planning into the architecture. In addition, HAC supports action abstraction, resource management, sensor integration, and is well suited to controlling...

We address the issue of learning and representing object grasp affordance models. We model grasp affordances with continuous probability density functions (grasp densities) which link object-relative grasp poses to their success probability. The underlying function representation is nonparametric and relies on kernel density estimation to provide a continuous model. Grasp densities are learned ...

Humans demonstrate remarkable capabilities to use a variety of grasp strategies for a given object depending on the context of the task often referred to as grasp affordances in dexterous hands. The interplay among the nervous system, the morphology of the human hand, and the properties of the object to demonstrate these advanced grasp affordance skills are not well understood yet. A key hypoth...

Interactions between the ventral premotor (PMv) and the primary motor cortex (M1) are crucial for transforming an object's geometrical properties, such as its size and shape, into a motor command suitable for grasp of the object. Recently, we showed that PMv interacts with M1 in a specific fashion, depending on the hand posture. However, the functional connectivity between PMv and M1 during the...

Robotic grasp detection task is still challenging, particularly for novel objects. With the recent advance of deep learning, there have been several works on detecting robotic grasp using neural networks. Typically, regression based grasp detection methods have outperformed classification based detection methods in computation complexity with excellent accuracy. However, classification based ro...

This paper presents a novel robot grasping planning approach that extracts grasp strategies (grasp type, and thumb placement and direction) from human demonstration and integrates them into the grasp planning procedure to generate a feasible grasp concerning the target object geometry and manipulation task. Our study results show that the grasp strategies of grasp type and thumb placement not o...

GRASP with path-relinking (GRASP+PR) is a metaheuristic for finding optimal or near-optimal solutions of combinatorial optimization problems. This paper proposes a new automatic parameter tuning procedure for GRASP+PR heuristics based on a biased random-key genetic algorithm (BRKGA). Given a GRASP+PR heuristic with n input parameters, the tuning procedure makes use of a BRKGA in a first phase t...

OBJECTIVE Recent studies have shown that dorsal premotor cortex (PMd), a cortical area in the dorsomedial grasp pathway, is involved in grasp movements. However, the neural ensemble firing property of PMd during grasp movements and the extent to which it can be used for grasp decoding are still unclear. APPROACH To address these issues, we used multielectrode arrays to record both spike and l...

A report of the General Robotics and Active Sensory Perception (GRASP) Laboratory. Comments University of Pennsylvania Institute for Research in Cognitive Science Technical Report No. IRCS-93-31. This other is available at ScholarlyCommons: http://repository.upenn.edu/ircs_reports/186

Tissue grasping is a crucial component of surgical procedures that involves simultaneously satisfying two competing criteria – maintaining grasp stability while avoiding damage due to excessive grip force. Automation of grasp force control thus requires a controller to apply grasping forces just sufficient to maintain grasp stability. This task is complicated by the serially connected dynamics ...