The introduction of mathematical modelling concepts to assess sensor performance is a necessary part of the graduate curriculum in remote sensing science. In this way, the concepts of sensor design and performance assessment under various imaging conditions can be best understood. In this paper, a graduate class project is described in which a static performance model was used to predict the minimum detectable temperature performance, the minimum resolvable temperature performance and the recognition performance of scanning and staring focal plane array (FPA) infrared imaging sensors with three different detector materials. Different seasonal, atmospheric, cold shield angle, target-to-background temperature contrast, bandpass and noise parameters were considered. Such studies are highly labour-intensive and (as in this case) may involve several hundred modelling runs. However, such projects provide a ‘real world’ experience for those who will design, use and analyse the data from imaging sensors. We present this paper as an approach to providing an in-depth understanding of sensor design concepts and limitations.