Failure detection in lighting system

The present invention relates to a method of detecting a failed luminaire and/or light sensor in a lighting system comprising M luminaires and N light sensors. The elements Dsl of an N×M transfer matrix D expresses how each luminaire l affect a light sensor s with respect to sensed light intensity, wherein s=1, 2, . . . N, and l=1, 2, . . . , M. The method comprises the steps of: setting the intensity of light emitted by the respective luminaires in the lighting system to a first intensity setting according to the respective elements of a first intensity vector I(1)(102), determining a first sensor intensity vector E(1) (104), setting the intensity of light emitted by the respective luminaires to at least a second intensity setting according to the respective elements of a second intensity vector I(2) (106), wherein at least one element of I(2) differs from a corresponding element or elements in determining a second sensor intensity vector E(2) (108), determining fault detection vectors from a difference between E(1) and D·I(1) and between E(2) and D·I(2), respectively (110), and identifying a failed luminaire and/or light sensor on basis of a difference between corresponding elements of the respective fault detection vectors (112). The present invention further relates to a failure detection system for detecting a failed luminaire and/or light sensor.