EAGER: Centralized Control of Large-Scale Distributed Sensor/Actuator Networks: Self-organizing Amorphous Façades

Abstract:

We wish to create office divider/facade systems that can adapt their visual appearance and privacy properties to your working needs, can be rearranged within minutes, and allow you to program them by gestures. For example, a wall could sense outside weather and its occupants' moods and adapt its color accordingly. Or, a wall could provide its user with privacy by becoming opaque and emitting white noise. Finally, individual elements could open themselves up to let air, noise, and little objects pass between zones. Key research challenges that we address along this way are

• aggregating and routing of sensor data in large-scale distributed systems,
• distributed control policies to control large numbers of actuators,
• distributed processing of sensing data,
• which are paramount in a large class of cyber-physical systems that deal with local sensing of physical parameters and large numbers of actuators.

Within this project, we have constructed a prototype of a modular wall system consisting of hexagonal elements that can change their opacity and color. We have investigated algorithms for disseminating and collecting information in such a system of many locally connected, miniature computing platforms (multi-cast routing) [5-6] and distributed processing of information using an application of gesture recognition. Gestures are recorded using capacitive sensors on each module, converted into a chain code, normalized and compared to an alphabet stored in a database. We studied both distributing computation and memory, which trades those resources against communication. We have also leveraged the hardware and software resulting from this research for two art installations [3,4], a soft sensing skin that can detect textures [2], and a novel composite material that can locally vary its stiffness [1].

Publications
[1] A. McEvoy, N. Farrow, N. Correll (2013): Toward computational smart materials with controllable stiffness. In: 19th Int. Conf. on Composite Materials, 2013.
[2] D. Hughes, N. Farrow, N. Correll (2013): Distributed Texture Identification and Localization in Artificial Skin.In: 2013 Int. Workshop on Soft Robotics and Morphological Computation, Monte Veritas, CH, 2013.
[3] N. Correll, N. Farrow, S. Ma (2013): Honey Comb: A platform for computational robotic materials. In: 7th Int. Conf. on Tangible, Embedded and Embodied Interaction (TEI), pp. 419-422, Barcelona, Spain, 2013.
[4] N. Correll, N. Farrow, K. Sugawara, M. Theodore (2013): The Swarm Wall: Toward Life's Uncanny Valley. In: K. Goldberg, H. Knight, P. Salvini (Ed.): IEEE Int. Conf. on Robotics and Automation, Workshop on Art and Robotics: Freud's Unheimlich and the Uncanny Valley, 2013.
[5] S. Ma, H. Hosseinmardi, N. Farrow, R. Han, N. Correll (2012): Establishing Multi-Cast Groups in Computational Robotic Materials. In: IEEE Int. Conf. on Cyber, Physical and Social Computing , Besancon, France, 2012.
[6] H. Hosseinmardi, R. Han, N. Correll (2012): Bloom Filter-Based Ad Hoc Multicast Communication in Cyber-Physical Systems and Computational Materials.The 7th Int. Conf. on Wireless Algorithms, Systems, and Applications (WASA 2012), 2012.