In the high-tech semiconductor industry floating air conveyors are used to transport and position silicon substrates. An interesting point for improvement is to limit the loss of pressurized air as conveyors are only effective at the position of the substrate. The goal of this research is to develop a local pneumatic valve, that is implemented underneath the surface of a floating air conveyor. Depending on the position of a substrate, the valve is actuated using a pressure signal. A conceptual design is conceived where a nozzle-flapper system builds up pressure when a wafer is floating nearby. This will actuate a relay with a circular waved membrane that snaps to its inverse shape, to open or close a channel. The proposed nozzle-wafer concept is modelled to investigate its characteristics. It is found that a desired nozzle back pressure can be achieved by setting the appropriate nozzle restriction and supply pressure. However, at higher pressures the fly-height will be influenced by the nozzles. Therefore, a minimal switching pressure for the relay is key. The arrangement of nozzles and relays in a conveyor is examined. In this research a black box is used for the valve in order for multiple relay concepts to be applicable. With the current function two inputs are needed to fully operate the relay. It is possible to make a grid where substrates move in one direction and an extra array of outlets are active around it, to make a smoother air film. No arrangement is found that can be used in different directions without leaving a single relay open when the substrate has passed. Simulations of the snap acting relay showed unreliable membrane behaviour. An adjustment is made in the design for the prototype. A design using a circular single curved membrane is introduced, because the models are consistent in converging to a solution of its snapping behaviour. A layered design approach is used for the prototype, where laser-cut PMMA plates are glued together. Plastic membranes are created with a vacuum form method. Initial testing of these membranes showed unexpected behaviour (e.g. flipping in other shapes and breaking). Also the pressure required to snap back to the starting position is considerable lower. A prototype is built using the only properly snapping membrane that is found. In a static experiment it proved that the relay functions as it is intended. This concludes the first iteration of the design process. A working snap acting membrane is fabricated and tested. Because of the behaviour of the other tested membranes it is taken into account that this prototype might not have a lifespan higher than switching a hundred times. Simulations showed that a nozzle-wafer system can provide a back pressure to switch a valve. However, this pressure must be minimal, in order that it does not affect the fly-height of the wafer. Lastly an alternative design is introduced. Using an elastic membrane with different size surfaces, lower pressures can close off air channels with a higher supply pressure. This combined by a nozzle-wafer conveyor with a resistor network underneath is recommended to investigate in the second iteration of the design process.