MIT Media Lab, E14-648 (Silverman Skyline Room)
Robotic vehicles walk on legs, roll on wheels, are pulled by tracks, pushed by propellers, lifted by wings, and steered by rudders. All of these systems share the common character of momentum transport across their surfaces. These existing approaches rely on bulk response among the fluids and solids. They are often not finely controllable and complex approaches suffer from manufacturing and practical operational challenges.
In contrast, Sun presents a study of a dynamic, programmable interface between the surface and its surrounding fluids. This research explores a synthetic hydrodynamic regime, using a programmable surface to dynamically alter the flow around an object. Recent advances in distributed computing and communications, actuator integration and batch fabrication, make it feasible to create intelligent active surfaces, with significant implications for improving energy efficiency, recovering energy, introducing novel form factors and control laws, and reducing noise signatures.
Sun's approach applies ideas from programmable matter to surfaces rather than volumes. The project is based on covering surfaces with large arrays of small cells that can each compute, communicate, and generate shear or normal forces. The basic element is a cell that can be joined in arrays to tile a surface, each containing a processor, connections for power and communications, and means to control the local wall velocity. The cell size is determined by the characteristic length scale of the flow field ranging from millimeters to centimeters to match the desired motion and fluidic system.
Because boundary layer effects are significant across fluid states from aerodynamics to hydrodynamics to rheology, the possible implications of active control of the boundary layer are correspondingly far reaching, with applications from transportation to energy generation to building air handling. This thesis presents a feasibility study, evaluating current manufacturing, processing, materials, and technologies capabilities to realize programmable surfaces.
Host/Chair: Neil Gershenfeld
John Deutch, John Bush, Alex Techet