TY - GEN
T1 - ORKA: An Ontology for Robotic Knowledge Acquisition
AU - Adamik, Mark
AU - Pernisch, Romana
AU - Tiddi, Ilaria
AU - Schlobach, Stefan
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.
PY - 2025
Y1 - 2025
N2 - Most autonomous agents operating in the real world use perception capabilities and reasoning mechanisms to acquire new knowledge of the environment, where perception capabilities include both the physical sensor devices and the software-based perception pipelines involved in the process. For autonomous agents to be able to adjust and reason over their own perception, knowledge of the sensors and the corresponding perception algorithms is required. We present the Ontology for Robotic Knowledge Acquisition (ORKA), that models the perception pipeline of a robotic agent by representing the sensory, algorithmic and measurement aspects of the perception process, thereby unifying the agent’s sensing with the characteristics of the environment and facilitating the grounding process. The ontology is based on the alignment between SSN and OBOE, linked to external databases as additional knowledge sources for robotic agents, populated with instances from two different robotic use-cases, and evaluated using competency questions and comparisons to related ontologies. A proof of concept use-case is presented to highlight the potential of the ontology.
AB - Most autonomous agents operating in the real world use perception capabilities and reasoning mechanisms to acquire new knowledge of the environment, where perception capabilities include both the physical sensor devices and the software-based perception pipelines involved in the process. For autonomous agents to be able to adjust and reason over their own perception, knowledge of the sensors and the corresponding perception algorithms is required. We present the Ontology for Robotic Knowledge Acquisition (ORKA), that models the perception pipeline of a robotic agent by representing the sensory, algorithmic and measurement aspects of the perception process, thereby unifying the agent’s sensing with the characteristics of the environment and facilitating the grounding process. The ontology is based on the alignment between SSN and OBOE, linked to external databases as additional knowledge sources for robotic agents, populated with instances from two different robotic use-cases, and evaluated using competency questions and comparisons to related ontologies. A proof of concept use-case is presented to highlight the potential of the ontology.
KW - Ontologies
KW - Robotic Knowledge Acquisition
KW - Robotic Perception
UR - http://www.scopus.com/inward/record.url?scp=85210882822&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85210882822&partnerID=8YFLogxK
U2 - 10.1007/978-3-031-77792-9_19
DO - 10.1007/978-3-031-77792-9_19
M3 - Conference contribution
AN - SCOPUS:85210882822
SN - 9783031777912
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 309
EP - 327
BT - Knowledge Engineering and Knowledge Management
A2 - Alam, Mehwish
A2 - Rospocher, Marco
A2 - van Erp, Marieke
A2 - Hollink, Laura
A2 - Gesese, Genet Asefa
PB - Springer Science and Business Media Deutschland GmbH
T2 - 24th International Conference on Knowledge Engineering and Knowledge Management, EKAW 2024
Y2 - 26 November 2024 through 28 November 2024
ER -