Combining declarative, procedural, and predictive knowledge to generate, execute, and optimize robot plans

Stulp, Freek; Beetz, Michael

doi:10.1016/j.robot.2008.08.011

Cited by 12 publications

(4 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Secondly, the physical agents need to deal with higher level knowledge for reasoning and planning in their environment. For this purpose, early approaches, such as the well-known Shakey robot, used the logical STRIPS representation (Fikes and Nilsson 1971), and several related symbolic representations (Hertzberg and Chatila 2008;Stulp and Beetz 2008) have been used later. To model both uncertainty and high-level action knowledge, we shall employ representations from statistical relational learning (SRL) (Getoor and Taskar 2007;De Raedt and Kersting 2008;De Raedt 2008) and probabilistic programming languages (PPLs) (De Raedt et al 2007), which combine expressive logical representations or programs with probabilistic reasoning and machine learning, Their use has recently been explored in robotics and their effectiveness has been shown in a kitchen scenario (Jain et al 2009).…”

Section: Introductionmentioning

confidence: 99%

Relational affordances for multiple-object manipulation

et al. 2017

View full text Add to dashboard Cite

The concept of affordances has been used in robotics to model action opportunities of a robot and as a basis for making decisions involving objects. Affordances capture the interdependencies between the objects and their properties, the executed actions on those objects, and the effects of those respective actions. However, existing affordance models cannot cope with multiple objects that may interact during action execution. Our approach is unique in that possesses the following four characteristics. First, our model employs recent advances in probabilistic programming to learn affordance models that take into account (spatial) relations between different objects, such as relative distances. Two-object interaction models are first learned from the robot interacting with the world in a behavioural exploration stage, and are then employed in worlds with an arbitrary number of objects. The model thus generalizes over both the number of and the particular objects used in the exploration stage, and it also effectively deals with uncertainty. Secondly, rather than using a (discrete) action repertoire, the actions are parametrised according to the motor capabilities of the robot, which allows to model and achieve goals at several levels of complexity. It also supports a two-arm parametrised action. Thirdly, the relational affordance model represents the state of the world using both discrete (action and object features) and continuous (effects) random variables. The effects follow a multivariate Gaussian distribution with the correlated discrete variables (actions and object properties). Fourthly,

show abstract

Section: Introductionmentioning

confidence: 99%

Relational affordances for multiple-object manipulation

et al. 2017

View full text Add to dashboard Cite

show abstract

“…New HIS approaches for distributed control may be inspired by the hybridization of reinforcement learning with evolutionary algorithms [99], but also by most classical mixtures of declarative, procedural knowledge and case based reasoning [163]. The enrichment of situational calculus in [54] with other Computational Intelligence tools, may enable this approach to be extended to very unstructured dynamic environments.…”

Section: Controlmentioning

confidence: 99%

On the potential contributions of hybrid intelligent approaches to Multicomponent Robotic System development

Duro

Graña

Lope

2010

Information Sciences

View full text Add to dashboard Cite

“…Robotics is a vast and active area seeking to develop mobile, physical agents capable of reasoning, learning and manipulating their environment. Early approaches such as in Shakey used logical representations such as STRIPS, and many more approaches use various other kinds of symbolic knowledge representation [2,9]. In addition to symbolic (or, semantic) methodologies, the physical aspect of robots requires dealing with various kinds of uncertainty typically not handled by symbolic formalisms.…”

Section: Introductionmentioning

confidence: 99%

Statistical Relational Learning of Object Affordances for Robotic Manipulation

Moldovan

Otterlo

Raedt

et al. 2014

Latest Advances in Inductive Logic Programming

View full text Add to dashboard Cite

We present initial results of an application of statistical relational learning using ProbLog to a robotic manipulation task modeled using affordances. Affordances encompass the action possibilities on an object, so previous works have presented models for just one object. However, in scenarios where there are multiple objects that interact between each other, it is very useful to consider the advantages of the statistical relational learning.

show abstract

Combining declarative, procedural, and predictive knowledge to generate, execute, and optimize robot plans

Cited by 12 publications

References 27 publications

Relational affordances for multiple-object manipulation

Relational affordances for multiple-object manipulation

On the potential contributions of hybrid intelligent approaches to Multicomponent Robotic System development

Statistical Relational Learning of Object Affordances for Robotic Manipulation

Contact Info

Product

Resources

About