NOAA estimates that the large local, state, and federal financial investments in nourishment projects to maintain existing beaches is returned ten-fold through tourism dollars alone. As sea level rises and population demands on our shorelines increase, some experts predict the socioeconomic and ecological pressures on the beaches and coastal planners will also rise. These projects are generally successful at temporarily restoring the beach width, however re-nourished sediment more quickly erodes as it has different characteristics than the "native" sediment. The PIs will develop and experimental program for these studies. Theories for the motion of sediment largely neglect any unsteady contributions from the passing waves or more local turbulent disturbance. The existing applied engineering models based on such principles have relatively low skill when predicting the onshore transport of sediment in moderate wave climates (e.g small wave periods of 3-5 s and wave heights of roughly 1 m) with weak mean flows where beaches rebuild and could significantly benefit from an improved understanding of nearshore sediment dynamics. The hypothesis the PI's seek to test is that in wave-driven coastal environments where the flow regime is unsteady and the sediment characteristics can be heterogenous, the incipient motion of a sediment bed results from a combination of the shear stress gradient and the horizontal pressure gradient. The PIs will test this hypothesis by developing a new wireless Lagrangian sensor, the Smart Sediment Grain. The specific objectives for this project are to develop a theoretical formulation for the incipient motion of sediment in response to random free surface gravity waves, extending Sleath's 1999 formulation; develop a Smart Sediment Grain (SSG) to measure the three-dimensional acceleration of individual motes (or grains); and finally observe the motion and transport of heterogeneous sediments (both natural and smart particles) through a comprehensive full-scale laboratory observation. Full-scale observations will be performed at the Ven Te Chow Laboratory at the University of Illinois (UIUC). Observations will be obtained for a range of natural grain classes (0.2 < d50 < 20.0 mm) and also for a range of lighter particles that will allow for a range of the full parameter space relevant to coastal environments. This effort represents a synergistic collaboration between a wireless technology expert and a sediment transport scientist. This effort is transformative in its application of state-of-the-art MEMs technology to resolve longstanding fundamental questions regarding the motion of intermittently mobile sediment beds exposed to free surface gravity waves. There exist wide-ranging broader impacts of this effort that include increasing the participation of underrepresented groups, transitioning high technology to society, entraining k-12 students into science and engineering, and increasing the predictive skill of coastal management models. This project will provide support for the Ph.D. studies of a minority female student. The PIs will participate in outreach programs at the Joan and James Leitzel Center at UNH and through Girls Inc, Eureka! Ongoing collaborations with TUDelft will provide an opportunity to model the successful Dutch methods for transitioning scientific findings to the larger engineering community.
The outcome of this project includes two types of motion-tracking systems for tracking underwater sediments: a miniature inertial sensor for in-situ sensing, and a camera-based technique for establishing ground truth. The inertial sensor system, called EcoSSG (smart sediment grain), is built by using two triaxial accelerometers controlled by a microcontroller with an on-chip data radio so that it can perform in-situ motion sensing (i.e., placed inside water to move with sand and gravel). It is the smallest gyro-free inertial measurement unit to date. The use of accelerometers without gyroscope enables it to achieve much lower power and small size. Each EcoSSG logs data to its on-board serial flash memory while being moved in water with sand and gravel. The data is later transferred wirelessly to a computer for postprocessing. In addition to inertial sensing, this project also developed advanced image-based motion tracking techniques. One was a fast and efficient motion deblurring technique, which constructs a clear latent image from a blurry source due to motion artifacts. The algorithm estimates the motion trajectory before performing the deblurring step. The patch-mosaic technique proposed by this project extracts the blur kernel from a subset of image patches and has been shown to be very effective and feasible to implement on mobile or embedded devices. Another advanced technique developed was to take advantage of inertial sensors (IMU) and depth sensors (Kinect) to further improve accurate estimation of the blur kernel. The result is the enhanced image for establishing ground truth for motion tracking. Together, the motion tracking techniques have been used in extensive sets of experiments in coastal environments and in laboratory settings. Experimental results confirmed the feasibility of the proposed in-situ measurement approach. Beyond the original application of tracking underwater sediments in coastal settings, the developed techniques have broad applicability to many science and engineering studies. One specific application of miniature in-situ sensing capability is infant monitoring, where the limb motion of infants can be a valuable indicator for preterm infant growth, cerebral palsy (if limbs exhibit cramped synchrony), as well as other health and fitness applications. The GF-IMU technique developed as part of this project is also applicable to a new generation of human-computer interface (HCI) devices similar to a finger-worn wireless mouse, except that it can work in 3 dimensions and recognize gestures in addition to mouse movements. The camera-based techniques also have great potentials for broader impact on digital photography in general. Because of the low complexity, it is feasible to implement on consumer-grade cameras, and it can potentially enable consumers to instantly deblur an image as soon as they finish taking it. This can not only enhance user experience but also enable better reconstruction of evidence from surveillance cameras.
