SEED GRANTS

To explore innovation in human-computer interfacing for virtual environments, this project focused on reconstructing environments from historical sites. This will eventually lead to the creation of networked virtual heritage sites connecting multiple UCs and other institutions. Its main objective was to create natural interfaces for collaborative virtual environments to facilitate computer interactions accessible for non-computer experts to program virtual environments. The researchers will also contribute to an ongoing effort to establish a center on Autonomous and Interactive Systems at UC Merced. Some notable accomplishments from the project include: developing software to run a visualization system at UC Merced, visualizing datasets from UC Davis through visualization software developed at UC Davis. Researchers have also developed a graphics framework to display a realistic visualization of the virtual heritage dataset of Villa di Livia. In the future, adjustments will need to be developed to connect with a 10-camera system being installed to track the user’s motions.

Power consumption inside buildings is one of the most critical factors when analyzing energy usage. The brand-new UC Merced campus represents a once-in-a-lifetime opportunity to create a full-scale test-bed for advanced energy management in campus buildings and systems. The new research university has made environmental performance, energy efficiency, and sustainability a campus-wide theme. Goals of this research include developing and deploying a sensor network system for real-time accurate estimation of person transitions among different part of the buildings when using public hallways. Researchers plan to utilize a 30 node smart camera test bed to be deployed in the Science and Engineering building. Through this test bed, they will collect extensive data to create a data-driven model that allows analysis and simulations. The project will enhance existing approaches to building design and operation with the results of its new measurement technologies.

Groundwater is often over-looked by the average Californian, but is an important—albeit overused—local source of water throughout the state. We currently use groundwater to close the gap between water demands and supplies from surface water and inter-basin transfers. The objective of the project was to create a prototype for an end-to-end Managed Aquifer Recharge Network (MAR-net) to help ensure a sustainable water supply for California. This project will design and test a scalable system for maintaining a continuous account of our groundwater resources. The project operated a prototypical MAR-net system in the Fresno Wastewater Treatment Plant from 10/2010 through 12/2011. The system included three vertical soil monitoring stations with moisture, temperature and salinity sensors, along with input flow metering, pressure transducers to measure pond depth, and a weather station. Researchers are currently developing a model to describe the data.

Real time motion capture and interactive 3D computer-generated environments are emerging as an integrated and powerful approach revolutionizing many applications. The potential for clinical and therapeutic applications is just beginning to be appreciated and explored in various areas of medicine, including telemedicine, stroke rehabilitation, pain management, and cognitive behavioral therapy. There are three main objectives for this research project: evaluate latest sensor technologies for use on virtual therapy applications, design an interactive prototype application for prescribed and personalized therapy rehabilitation, and to evaluate novel solutions for tracking finger, hand, and facial skin deformations for rehabilitation of hand motor control and facial expressions after burn injuries. This research resulted in a working prototype system for modeling upper-body physical therapy exercise motions and for automated monitoring of patients following exercises demonstrated by an animated 3D character. A provisional patent application about the main techniques was developed and two poster papers were presented and published at the IEEE VR conference.

Sustainable buildings represent a major paradigm shift from current practice. According to the U.S. Department of Energy, buildings accounted for near 40% of US primary energy consumption in 2006, 75% of which is electrical energy. At the same time, energy and performance ratings are becoming more actively pursued and new methods are being sought for performance certification based on operational measurements. This defines a need to combine state of the art sensor network technology with novel processing and control algorithms in order to determine the distribution of people inside buildings and actuate the HVAC system using this information. Researchers plan to use an existing wireless sensor test bed of micro-cameras deployed on the SE1 Building at UC Merced in order to measure real-time occupancy. Further, they plan to develop a system to take the real time occupancy data from the testbed and interface with the EMS system in order to actuate part of the building. They plan to run control experiments obtaining quantitative and qualitative data of energy savings and levels of thermal comfort for the users.

The ability to move independently is a critical component of a person’s quality of life. For many people, however, independent mobility is hampered by one or more forms of disability: motor, cognitive, or sensorial. For example, blind persons experience difficulty moving and finding their way in a previously unexplored environment, due to lack of visual access to the features that sighted persons use for self-orientation. Likewise, some wheelchair users may find it challenging or impossible to safely and reliably control their wheelchair due to poor upper limbs control or to cognitive impairment. The main goal of this project is the development of computer vision techniques that can be used to assist both categories of users mentioned above: persons who, due to visual impairment, experience difficulty self-orienting, and persons with mobility impairment who are unable to independently control the trajectory of their motorized wheelchair. Specifically, researchers will develop technology that promotes semantic spatial awareness in manmade environments by means of direct access to printed textual information.

Urbanized regions are responsible for a disproportionately large percentage (30-40%) of global anthropogenic greenhouse gas emissions. While a growing number of studies have addressed urban and suburban CO2 fluxes, current experimental techniques are not robust with respect to policy or scientific objectives. Thus, researchers explored a novel technique for surface flux partitioning based on coupled atmospheric models and continuous sensor-network observation of ambient CO2 and carbonyl sulfide. The goal of the project was to create proof of concept results and a science plan for developing, deploying, and assessing a new COS sensor and atmospheric modeling strategy for the application of greenhouse gas emissions measurement at the city scale. Researchers initiated the implementation of a unique top-down constraint on an urban GHG budget using the simultaneous analysis of atmospheric CO2 and carbonyl sulfide.

One of the key challenges in archaeology is successfully communicating data, even within the scientific community. Although there has been exponential growth in the production of 3D content for archaeology, very few are accessible. As a result, a very small percentage of information that is produced is well communicated. This has an adverse impact on the interpretation process, where the process remains an isolated experience without feedback. This project can create the necessary infrastructure to promote discussion and improve interpretations in real time. The primary goal of this project is to create a network of virtual collaborative environments across UC campuses to develop novel techniques for collaborative research and learning in archaeology. These environments allow users to interact and learn in rich 3D virtual spaces, where they can exchange data and information. Through these environments, history, archaeology and other scholars can collaboratively interpret reconstructed heritage artifacts, sites, and landscapes. This project will extensively integrate different immersive systems and 3D web virtual environments.

Every day, robotic technologies save human lives in a variety of situations. Robots can go where it is unsafe for human beings in first response situations, and they can be fully controlled by trained operators. Usually, the robot operator experiences the situation being assessed only through a user interface combining a set of perceptual cues provided by the robot sensors. Moreover, the operator interacts with the robot using control devices similar to those commonly used to control virtual reality artifacts, like video game characters or synthetic avatars operating in cyberspace. Thanks to a synergy between the robotics group at UC Merced and the Merced Sheriff Office SWAT team, the goal of this project is to explore how this interplay between virtual reality and robotics can lead to the development of better training tools for first responders. A unique feature of this project is the partnership between end-users already using tactical robots and a university research team with multi-year experience in the field of Urban Search and Rescue (USAR) robots.

This proposal seeks to enhance machine understanding of big image and video data sets by including human computational units as an element inside larger computational systems. Cameras generate by far the most data in the world, and even small organizations can easily deploy cameras to monitor social and environmental problems. Thus, this project has four goals: characterize and benchmark human computation co-processors (HPUS) platforms, propose instructions and algorithms, implement instructions and algorithms, and evaluate work. Researchers seek to build an automated test suite of small tasks that can be run on a wide set of HPU platforms, providing a snapshot of the accuracy, latency, and reliability of each platform. Researchers need to evaluate their work, which will be done using applications already under development that have previously suffered from insufficient robustness of existing computer vision libraries. Lastly, researchers will quantify performance against existing solutions.

In recent years, multiple works have been published showing how wireless sensor networks can be utilized to reduce the energy consumed by buildings. However, within the wireless sensor network community, there has been little research attempted to improve the quality of service for users. Instead, the aim has been simply to maintain, or in some cases, potentially decrease the quality of service in order to achieve greater efficiency. The goal of this research is to increase the conditioning effectiveness of the HVAC system by incorporating participatory sensing and making adjustments to the building management systems in real time. Part of this goal is to show how participatory sensing is an effective tool for augmenting and maintaining the building management systems.

Researchers at UC Merced designed and deployed 2 advanced solar observatories at the UC Merced and UC Davis campuses. In order to implement an effective policy for solar power connectivity, the ability to forecast both short- and long-term solar availability is paramount to utility companies and large-scale power generators. CITRIS researchers utilized SIMI, a new concept that combines state-of-the-art satellite and radar image processing with real-time data from ground solar stations to forecast solar power availability for the various atmospheric conditions found around California. The web-based, near real-time information database generated by SIMI is providing a direct estimate of both current and prospective power availability for solar-based technologies in the state of California, as well as facilitating the deployment and connectivity of solar power plants to the electric grid. The data will be used in decision-making regarding installation, incentives and policies for solar technologies through the consideration of geographical and financial constraints. One of the expected results from this initiative is a complete web-based GIS (Geographic Information System) that facilitates the promotion and management of solar energy utilization in California.

Sepsis is a severe medical condition with high hospitalization cost and high mortality rates, with more than half of the patients with septic shock dying within 30 days. While scoring tests and alert systems are in place for sepsis patients, researchers currently lack a data-based tool to guide clinicians on early prognosis and effective treatment of sepsis. Researchers propose to create an adaptive, data-driven, probabilistic clinician decision support system (CDSS) that will employ state of the art machine learning tools to mine the data rich Electronic Medical Records of the UC Davis Health System and can be readily deployed throughout hospitals in CA and around the country. Specifically, the project will be organized in three main phases: data mining, system design and implementation, assessment and refinement.

Respiratory epithelial cells, which form the lining of the airway system, provide our bodies its first line of defense against many harmful inhaled agents including dusts, air pollutants, chemicals and microbes, in addition to many biological and chemical agents. CITRIS researchers aimed to design and develop small-scale biosensors to incorporate airway epithelial cells to detect airborne pollutants through microfluidic devices. These devices detect harmful airborne agents, which, when integrated into a wireless system, would enable a detection system linked into a real-time communications network. These sensor communications networks could provide a much needed early warning system for sensitive areas, such as in large population centers or areas with polluted air. This project also represents a unique opportunity to apply sensor network technology to monitor the air quality in the San Joaquin Valley, which ranks amongst the most polluted air basins in the US and has been designated by EPA as a “nonattainment” area. However, biosensors for nanoparticles are still under-developed. This project aims to develop a highly sensitive and potential portable whole cell-based biosensor with the capability to detect the presence of a wide-variety of nanoparticles for air quality control.

The Center for Sierra Nevada Water Information System (CSNWIS) project resulted in a web portal which provides Sierra Nevada satellite-snow-cover information, wireless-sensor-network data, and analyses which are useful for water scientists, engineers, regulators, and managers. CSNWIS has created snow-covered-area datasets for 2000-2011 years which are currently available for viewing and download. Basin-scale analyses of snow products by elevation are also available through the web portal. Using a modular, scalable design, the CSNWIS provides a platform for archiving and serving water information. CSNWIS will continue to expand with more data sets to provide information to users with research and operational forecasting interests. In the future, projects may be expanded through collaboration with the California Resources Agency, hydroelectric companies, and local water providers.

Transdermal patches are the least invasive of available drug delivery techniques. Since their approval for clinical use in transdermal patches have become ideal carriers for a variety of drugs. Although transdermal patches are very promising, even wider application of this drug-delivery technique is still hindered by two main factors: the limited permeability of the human skin and the absence of a cheap, reliable and versatile patch that can release well-controlled dosages of a variety of drugs at precise time intervals over a long period. This project aims to design a smart controllable membrane, which combines the simplicity of a membrane based patch with the precise drug delivery of a complex micro-machine based system. This has the potential to decrease the cost and increase the applicability of transdermal delivery. The aim of the research is to investigate whether a modeling approach encompassing two complementary simulation techniques in an integrated framework can be used to design a smart membrane. Specifically, the goal of the proposed research is to investigate how porous membrane surface features could be rationally designed for precise flow control.

Large and complex HVAC systems are naturally difficult to monitor. Indeed, one of the outcomes of this research was the development of a top-down strategy for fault detection, which discovered several faults in the Science and Engineering building. The objective of this project is to develop methods for effective monitoring of system conditions and fault detection of large and complex HVAC systems. This project contributes to the development of intelligent control, monitoring, and fault detection technologies that are integral part of energy-efficient carbon-neutral buildings. The key outcome of the project includes methods for spatial and temporal partition of large complex HVAC systems, development of pre-processing algorithms to discover new features in the monitored data set, selection of effective methods for HVAC monitoring and fault detection, and finally establishment of thresholds for various faults.

Intelligent water information systems have the potential to provide precise measurement of snowpack and its condition from day to day and particularly at critical points in the annual cycle. This information can be translated into more accurate flow predications, presenting managers with the opportunity to improve the operation of water infrastructure, including power generation, water supply, flood protection, and extend the functional life of existing infrastructure. The goal of this project is to collect and analyze data on the economic uses of water, understand the value of accuracy and timeliness of this information, and develop an improved economic model that can assess the value of better information concerning the release of water from watersheds above the river and reservoir system of California. Using sensor-based water management technologies, the researchers will select an initial case study for the work, and include recommendations on how the approach could be adapted.