IoT environments consist of heterogeneous devices, equipped with actuators and sensors. HVAC systems are an example of an IoT environments, measuring real world variables through sensors and acting upon the read measurements through actuators. To bring the sensors and actuators, of these environments, into context with each other and the real world, they can be modelled using ontologies. The resulting context models define relations between the sensors, actuators, and the real-world environment. These IoT environments are, however, prone to errors. False sensor values may cause incorrectly triggered actuators and an overall incorrect behaviour of the application. To avoid these issues, a validation of the measurements must be performed, to filter out outliers. In this thesis, an architecture is proposed, that enables the definition of validation methods within the context model. By running queries against the application, time series measurements can be extracted without the need for any additional information. The proposed approach, stores live measurements, as well as the context of the IoT devices and their hierarchical structure, in a real-world context, inside a single context model. This enables, validation across different devices and systems. Furthermore, instructions are provided in the thesis, describing how the architecture can be replicated, using different validation methods and platforms.