Opportunistic sensing becomes a competitive sensing paradigm nowadays. Instead of pre-deploying application-specific sensors, it makes use of sensors that just happen to be available to accomplish its sensing goal. In the opportunistic sensing paradigm, the sensors that can be utilized by a given application in a given time are unpredictable. This brings the Semantic-Web based situation inference approach, which is widely adopted in situation-aware applications, a major challenge, i.e., how to handle uncertainty of the availability and confidence of the sensing data. Although extending standard semantic-web languages may enable the situation inference to be compatible with the uncertainty, it also brings extra complexity to the languages and makes them hard to be learned. Unlike the existing works, this paper developed a situation inference tool, named OPSitu, which enables the situation inference rules to be written in the well accepted standard languages such as OWL and SWRL even under opportunistic sensing paradigm. An experiment is also described to demonstrate the validity of OPSitu.