SHF: Small: Analytical Modeling and Design Methodology for Large-scale Computational Systems Employing Flexible Electronics for Extensive Physical Interfacing

点击翻译按钮获取中文摘要

Electronics has enabled substantial computational capabilities in small-scale devices; this means that signals from physical systems can be presented to these devices to derive high-value outputs. The challenge is that although immense computational capacity can be realized in embedded devices, the ability to acquire a large number of signals that are distributed across a physical system, and to then communicate information over the associated distances, remains disproportionately small. Technologies are emerging, however, that enable transformational possibilities for creating communication channels and large-scale physical interfaces to electronics. Large-area electronics is a technology that enables the fabrication of interconnects as well as expansive arrays of diverse sensors on flexible, low-cost sheets. When combined with high-performance silicon integrated circuits (ICs), this technology can lead to systems where computation can be applied to physical signals on a much larger scale than that possible today. The resulting design space for the systems covers two technology domains. To create efficient and scalable systems, the platform components and hardware architectures must be analyzed rigorously, and methodologies for understanding and optimizing design trade-offs must be developed. The objective of this research is to analytically model the platform components and architectures for sensing and communication, and then to synthesize these into system design methodologies. The platform architectures include interfaces for digital and analog signaling, control circuits for sensing, and communication networks scalable to many nodes. The methodologies span analysis at the device, architecture, and system-protocol levels. Some of the compelling applications that require large-scale interfacing of electronics with physical systems include detection of early-stage structural degradation in bridges and buildings through centimeter-resolution strain sensing, interactive surfaces for visually-rich computing via high-resolution displays and input sensors, etc. Large-area electronics and high-performance ICs, used synergistically, have the potential to enable such applications. By developing analytical models and methodologies for system design, this research aims to unite the efforts from both the technology-development and computer-systems domains. The outcomes of this study will help coordinate and focus research and engineering efforts in these areas towards the creation of optimal systems. This research will also engage a new generation of engineers, particularly from underrepresented groups, through the unique opportunity to develop systematic principles for the design of computing systems that interact extensively with physical systems. The broadened focus and new form-factors possible for computing systems will be illustrated to students through undergraduate projects and special-topic courses, as well as through Princeton outreach programs such as the Science and Engineering Expo for middle-school students, the Materials Camp for teachers, and the Princeton University Materials Academy for under-represented high-school students.