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.