Abstract: A multiple dimension vector-based occupancy sensor. One or more sensors (for example, ultrasonic and/or infrared sensors) are used to produce one or more occupancy estimator vectors. The individual dimensions of the one or more occupancy estimator vectors are combined using a fusion based detection algorithm to produce a combined estimator signal. The combined estimator is then used to switch electrical loads when the combined estimator signal is above a predetermined threshold. The result is an occupancy sensor that has a very high probability of occupancy detection and a low probability of false tripping.