Current research in intelligent systems investigates their deployment in dynamic and complex environments. Such systems require the capability to be aware of their operating environment and to process effectively sensory information from multiple sensory sources. The abilities observed in the animal kingdom to process sensory information in varying conditions, from many different sensory sources, is an inspiration for intelligent systems research. Sensory processing in the mammalian brain involves thousands of neurons in cortical columns, with extensive interconnect. However it is known that interconnections between neurons and thus the source of spiking activity within these biological columns is locally based. Cortical columns are also stimulated by connections from related areas within the brain which are dedicated to the processing of alternative sensory stimuli. This paper reports on an approach to emulate biological sensory fusion, based on Spiking Neural Networks (SNN) and Liquid State Machines (LSM), and is assessed in experiments involving the control of a mobile robot in a reactive manner. The results show that the sensory processing provided by the Liquid State Machine enables the reactive control of the robot within its environment.