Manipulation primitives: A paradigm for abstraction and execution of grasping and manipulation tasks

Felip, Javier; Laaksonen, Janne; Morales, Antonio; Kyrki, Ville

doi:10.1016/j.robot.2012.11.010

Cited by 25 publications

(10 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Grasping an unknown object is performed by executing a top grasp, closing fingers around the centroid of the point cloud of the object to grasp, similar to e.g. [9].…”

Section: Methodsmentioning

confidence: 99%

Robotic manipulation of multiple objects as a POMDP

Pajarinen

Kyrki

2017

Artificial Intelligence

View full text Add to dashboard Cite

This paper investigates manipulation of multiple unknown objects in a crowded environment. Because of incomplete knowledge due to unknown objects and occlusions in visual observations, object observations are imperfect and action success is uncertain, making planning challenging. We model the problem as a partially observable Markov decision process (POMDP), which allows a general reward based optimization objective and takes uncertainty in temporal evolution and partial observations into account. In addition to occlusion dependent observation and action success probabilities, our POMDP model also automatically adapts object specific action success probabilities. To cope with the changing system dynamics and performance constraints, we present a new online POMDP method based on particle filtering that produces compact policies. The approach is validated both in simulation and in physical experiments in a scenario of moving dirty dishes into a dishwasher. The results indicate that: 1) a greedy heuristic manipulation approach is not sufficient, multi-object manipulation requires multi-step POMDP planning, and 2) on-line planning is beneficial since it allows the adaptation of the system dynamics model based on actual experience.

show abstract

“…Grasping an unknown object is performed by executing a top grasp, closing fingers around the centroid of the point cloud of the object to grasp, similar to e.g. [9].…”

Section: Methodsmentioning

confidence: 99%

Robotic manipulation of multiple objects as a POMDP

Pajarinen

Kyrki

2017

Artificial Intelligence

View full text Add to dashboard Cite

show abstract

“…For robot planning, these tasks have been subdivided into more general manipulation primitives or elemental actions; e.g. move, approach, transport, place, push, slide, grasp, release [22], [23]. A second way to describe tasks is from the human controller point of view.…”

Section: Selection Of a General Set Of Tasksmentioning

confidence: 99%

A task analysis approach to quantify bottlenecks in task completion time of telemanipulated maintenance

Boessenkool¹,

Wildenbeest

Heemskerk³

et al. 2018

Fusion Engineering and Design

View full text Add to dashboard Cite

Telemanipulation techniques allow for human-in-the-loop assembly and maintenance tasks in otherwise inaccessible environments. Although it comes with limitations in achieved performance -required strict operator selection and extensive training are widely encountered -there is very little quantitative insight in the exact problems operators encounter during task execution. This paper provides a novel hierarchical task analysis approach to identify the most time-consuming subtask elements and to quantify the potential room for performance improvement during telemanipulated maintenance tasks. The approach is illustrated with a human factors case study in which 5 subjects performed six generic maintenance tasks, using a six degree of freedom master device connected to a simulated task environment. Overall it can be concluded that the proposed Three Phased Task Analysis is a useful tool to guide improvements since it is able to relate high-level problems (e.g. large variability) to behaviour on lower task-levels. For the case study, the largest potential for improvement was found for specific subtasks characterized by complex contact transitions and precise control of tool orientation, and in the reduction of variation of the task execution.

show abstract

“…Grasping techniques are classified depending on whether the target objects are previoulsy known, unknown or similar to a familiar category. Felip et al (2013) propose a functional organization of grasping which relies on the concept of manipulation primitive. For what concerns the planning of grasping actions through the use of visual information, a number of simplifications have commonly been used in robotics.…”

Section: Vision-based Robotic Grasping a Brief Outlinementioning

confidence: 99%