Welcome to EverybodyWiki ! Nuvola apps kgpg.png Log in or create an account to improve, watchlist or create an article like a company page or a bio (yours ?)...

Open Everyday Activity Science and Engineering

From EverybodyWiki Bios & Wiki

openEASE (Open Everyday Activity Science and Engineering) is a web-based knowledge engine for robots, which stores, maintains and allows reasoning about robot and human experience data of everyday activities [1]. This system is developed at the Institute for Artificial Intelligence at the University of Bremen, Germany.

Infrastructure[edit]

openEASE uses the SWI-Prolog RDF infrastructure from KnowRob knowledge base, which makes use of procedural attachments to OWL classes and properties for integrating external reasoners. The same way KnowRon, openEASE can be queried using the Robot Operating System and Docker (software).

Representations[edit]

openEASE contains semantically annotated data of actions, like manipulation, navigation, etc. in a episodic memory. It also includes the information about the environment the agent, human or robotic, is acting in, objects, and the behavior it generates. The Knowledge representation and reasoning in this case is presented as formal statements using pre-defined templates.

Episodic memories[edit]

Episodic memory is related to the experience information, which is organized temporally and spatially, alongside combined with context information. The episodic memories managed by openEASE include representations that combine symbolic plan events (high-level) with subsymbolic sensor data (low-level), and time-indexed of the events[2]. An episodic memory is understood as a recording that the agent makes of the ongoing activity, which includes very detailed information about the actions, motions, their purposes, effects and the behavior they generate, it also includes the images captured during execution, etc.[3]

Usage[edit]

openEASE is able to answer questions about the episodes by using queries based on Prolog and inference tools based on SRDL. Both operate on the robot model and are able to check dependencies of action descriptions. The system allows reasoning about the data and retrieves requested information based on semantic queries[4] [5], see the example below:

Question: Show a contact event and reason which one is a collider
Query:
 ep_inst(EpInst), entity(E, [an, event, [type, knowrob_u:'TouchingSituation']]), 
 occurs(E, [S, _]), comp_contact_roles(EpInst, E,CollisionEvent), 
 show_world_state(EpInst, S).

openEASE can be used by humans while using a browser-based query and visualization interface, but also remotely by robots via a WebSocket API[6]. For this reason, openEASE can be seen as a Cloud robotics infrastructure, as any robotic platform is able to query information regarding consequences of actions. It also provides massive storage and computation capacities. It has been cited in Cloud robotics literature by Koubaa et al. work in cloud-based aerial vehicles tracking[7] and the work from Saxena et al. in the construction of knowledge engines[8]. OpenEase has also being used by other research groups, as the University of Tokyo, which episodes can be found in the website[9]. The publication of this work has been nominated to the best paper award.

References[edit]

  1. M. Tenorth, J. Winkler, D. Bessler, M. Beetz (2015) Open-EASE: A Cloud-Based Knowledge Service for Autonomous Learning. KI - Künstliche Intelligenz 29(4):407-411. ISSBN 1610-1987. doi: 10.1007/s13218-015-0364-1.
  2. J. Winkler, M. Tenorth, A. K. Bozcuoglu, and M. Beetz. CRAMm– Memories for Robots Performing Everyday Manipulation Activities. Advances in Cognitive Systems, 3:47–66, 2014.
  3. J. Bateman, M. Beetz, D. Bessler, A. Kaan Bozcuoglu, M. Pomarlan (2017) Heterogeneous Ontologies and Hybrid Reasoning for Service Robotics: The EASE Framework. In Third Iberian Robotics Conference, Sevilla, Spain. (Accepted for publication).
  4. Georg Bartels, Daniel Bessler, Michael Beetz, Moritz Tenorth, Jan Winkler (2015) How to Use OpenEASE: An Online Knowledge Processing System for Robots and Robotics Researchers. In Proceedings of the 14th International Conference on Autonomous Agents and Multiagent Systems (AAMAS), Istanbul, Turkey.
  5. J. Wielemaker, W. Beek, M. Hildebrand and J. Van Ossenbruggen (2016) ClioPatria: A SWI-Prolog Infrastructure for the Semantic Web. Semantic Web, 7(5), 529-541. doi: 10.3233/SW-150191
  6. G. Bartels, D. Bessler, M. Beetz, M. Tenorth, J. Winkler (2015) How to Use OpenEASE: An Online Knowledge Processing System for Robots and Robotics Researchers. In Proceedings of the 14th International Conference on Autonomous Agents and Multiagent Systems (AAMAS), Istanbul, Turkey.
  7. A. Koubaa and B. Qureshi (2018) DroneTrack: Cloud-Based Real-Time Object Tracking Using Unmanned Aerial Vehicles Over the Internet, in IEEE Access, vol. 6, pp. 13810-13824. doi: 10.1109/ACCESS.2018.2811762
  8. Saxena, Ashutosh; Jain, Ashesh; Sener, Ozan; Jami, Aditya; Misra, Dipendra K.; Koppula, Hema S.(2014) RoboBrain: Large-Scale Knowledge Engine for Robots. ArXiv e-prints.
  9. A. Bozcuoglu, K. Gayane, Y. Furuta, S. Stelter, M. Beetz, K. Okada, and M. Inaba (2018) The Exchange of Knowledge using Cloud Robotics. Robotics and Automation Letters 3(2): 1072-1079. doi: 10.1109/LRA.2018.2794626

External links[edit]


This article "Open Everyday Activity Science and Engineering" is from Wikipedia. The list of its authors can be seen in its historical and/or the page Edithistory:Open Everyday Activity Science and Engineering. Articles copied from Draft Namespace on Wikipedia could be seen on the Draft Namespace of Wikipedia and not main one.


Farm-Fresh comment add.png You have to Sign in or create an account to comment this article !