Characterizing Mechanical Properties of Micro-Springs for MEMS Devices (2010)
Student: Hang An
Faculty Mentor: Prof. Khalil Najafi
entor: Dr. Becky Peterson, Research Scientist
Graduate Student Mentor: Seow Yuen Yee
Acoustic resonators are used in a variety of applications, including ultrasound sensors, loudspeakers, and a variety of other sensors and actuators. These devices have been miniaturized using micromachining technologies and have been recently used in our research group in applications for gas micro pumps and propulsion of micro air vehicles. The undergraduate student will be supporting the research of a graduate student to model the thrust, velocity, and pressure for a micro-acoustic resonator (also referred to as micro machined acoustic ejector) that is being designed for generating thrust and propulsion in a micromachined flying silicon wafer. The undergraduate student will be investing time in optimizing the performance of the micro-acoustic ejector for maximum thrust and velocity using commercial simulation software such as COMSOL Multiphysics. Candidates should have some background in mechanics and/or fluid dynamics, and preferably would have some experience modeling with finite element analysis software such as COMSOL Multiphysics.
Development and Characterization of Ink-jet Printing for Wafer Bonding/sealing of Integrated Microsystems (2010)
Student: Yemin Tang
Faculty Mentor: Prof. Khalil Najafi
Mentor: Dr. Becky Peterson, Research Scientist
Graduate Student Mentor: Ali Besharatian
Robust, low-temperature bonding and sealing methods are needed to package many micro electromechanical systems (MEMS) such as micro-pumps, gyroscopes, and pressure gauges. In this project, the student will work closely with graduate students to develop new sealing technologies for a miniature gas micro pump. Specifically, the student will use equipment in the Lurie Nanofabrication Facility to deposit sealant materials such as epoxies (glues) or other soft materials onto substrates such as silicon or glass using a specialized ink-jet printing machine, in order to bond two substrates together. S/he will be responsible for identifying appropriate sealants, developing the printing technique for each sealant, and assessing the impact of different substrate preparations. The student will test the effectiveness of the new sealing methods using a variety of tests including non-destructive inspection, bond shear strength measurement, leak/permeability testing, and thermal durability testing. Candidates should have some knowledge of electronic materials (e.g. semiconductors) or materials science. A background in micro-/nano-fabrication is preferred but not required.
Eliminating Concurrency Bugs with Control Engineering: Control Logic Synthesis (2010)
Faculty Mentor: Stephane Lafortune
The "Gadara'' project at the University of Michigan involves a multi-disciplinary team from the EECS Department. The goal of this project is to use techniques from control engineering to control the execution of concurrent programs in order to avoid deadlocks at run-time. Gadara consists of two parts. At compile time, a model of the program is created to identify potential deadlocks. At run-time, control logic embedded in the program throttles lock acquisition of threads to ensure that deadlocks are avoided. Our team has built a first prototype of Gadara that implements this approach. See the Computer Magazine of IEEE, pages 52-60 in the December 2009 issue, for a general description of Gadara.
This summer project is concerned with improving the control logic synthesis part of Gadara. This control logic is synthesized from the model of the program generated at compile time; the synthesis method exploits the theory of supervision based on place invariants for Petri nets and is formalized in the framework of linear algebra. The undergraduate student intern will work closely with one of the graduate students in the Gadara team to implement a new set of control logic synthesis algorithms, developed by our team, in Matlab. The student will then test these algorithms on sample programs and analyze their performance. The results of this work will be used to determine which algorithms to include in the current Gadara prototype.
Energy Harvesting from Piezoelectric Devices (2010)
Student: Miran Al-Haideri
Faculty Mentor: David Wentzloff
Piezoelectric crystals are small devices that convert mechanical stress into electric potential, which can ultimately be used to convert oscillating mechanical motion into a continuous power source. This project involves utilizing piezoelectric technology as an energy harvester for battery replacement in portable applications or as an alternative energy source. Piezoelectric crystals will be characterized with different mechanical excitations to quantify their theoretical maximum efficiencies. Power electronic circuits will be designed that interface to the piezoelectric devices to convert the energy into a more useful form for storage.
Energy-efficient VLSI Design of Massively Parallel Signal Processing Systems (2010)
Student: Yaoyu Tao
Faculty Mentor: Zhengya Zhang
Objectives: Advances in VLSI technology allow sophisticated signal processing to be performed with a smaller cost. However, an increasingly tighter energy envelope for mobile and sensing devices place an upper bound on the achievable computational intensity. In this project we consider the co-optimization of a class of computationally-intense signal processing algorithms and highly parallel hardware architectures for new opportunities in energy reduction
Methodology: We will focus on the motion estimation algorithm in video compression. Interesting operating characteristics can be identified for different parts of the algorithm as well as their mappings in datapaths and memories. Low-energy digital circuit techniques and memory paradigm will be tailored to the operating characteristic of each datapath and storage. The tradeoff between energy and performance can be exploited in maximizing the computational intensity.
Undergraduate Role: The student's research will be carried out under the supervision of a graduate student and a faculty mentor. The student will be expected to study literature to gain the background understanding of the application, followed by building the hardware prototype in Simulink. As the student becomes more familiar with the designs, she/he will perform investigations through simulations and real-time emulations on FPGA to explore the design space. The final design will be systematically optimized based on tradeoff analysis.
Helical CT Image Reconstruction (2010)
Student: Gregory Handy
Faculty Mentor: Jeff Fessler
Prof. Fessler's research group is developing iterative algorithms for reconstructing X-ray CT images from the 3D data recorded by CT scanners. The overall goal of this research is to develop sophisticated algorithms that can provide improved image quality at lower X-ray doses. This project will involve using mathematical tools to develop computationally efficient approximations of the image noise for these image reconstruction methods and to verify those approximations using computer simulations. Although the ultimate application of this work is biomedical imaging, and the student(s) involved in the project will gain exposure to medical imaging problems, the primary focus of this particular project is very mathematical and best suited for a student with a strong mathematical background.
A journal paper that describes some background for this project is online. The mathematics in this paper are quite challenging; the reference is included primarily to provide a sense of the type of research.
Multiple Target Tracking with Single Minimally Calibrated Camera (2010)
Student: Yan Qi
Faculty Mentor: Silvio Savarese
The ability to interpret the semantic of objects, their geometric attributes as well as their spatial and functional relationships within complex environments is essential for an intelligent visual system. In visual recognition, the problem of categorizing generic objects is a highly challenging one. Single objects vary in appearances and shapes under various photometric (e.g. illumination) and geometric (e.g. scale, view point, occlusion, etc.) transformations. In order to address these challenges, researchers in the Vision Lab at the University of Michigan are developing cutting edge technology in visual recognition. Our goals are to: i) introduce novel representations for describing rigid and non-rigid object categories. ii) Design methodologies for learning multi-view models where training data is provided in an unorganized fashion (e.g, from the Internet); iii) design algorithms for accurate object detection and view point estimation from either images or video sequences. Be part of the team and get involved in one of our projects! Tools developed in this project are critical in a large number of applications such as autonomous vehicle navigation, robot sensing and manipulation, mobile vision, post-production movie editing, image database indexing, and human-computer interface. Our technology can play a fundamental role in designing systems that can help people with reduced functional capabilities due to aging or disability toward the goal of improving and sustaining the quality of life for all people.
Prototype Development for Next-Generation Ultra-Low Power Wireless Electronic Devices (2010)
Faculty Mentor: Dennis Sylvester
Low power research at the University of Michigan has resulted in the most energy-efficient microprocessor ever reported. This technology enables multi-year or multi-decade battery life in wireless devices in a paper-thin form factor, creating new opportunities in a wide range of markets, including medical sensors, energy monitors in buildings/homes, and smart credit cards. It is the goal of this project to use the energy-efficient microprocessor to develop several hardware prototypes that demonstrate low power operation in multiple markets. Each prototype will consist of a printed circuit board, our low power microprocessor, a host of peripherals (displays, batteries, energy scavengers, etc.), and software to run on the microprocessor.
We are seeking assistance in developing both hardware and software for these prototypes. Interested candidates should have an understanding of basic electronic circuits and programming and would preferably have experience in printed circuit board design, semiconductors, and chip design.
Reliability of Organic Solar Cells (2010)
Student: Sheila Raja
Faculty Mentor: Stephen Forrest
Objectives: To understand the limitations to the operating lifetime of solar cells made from organic thin film materials
Methodology: To develop methods for accelerating the degradation of cells made in our laboratory. This will involve developing a package for the solar cells that allows for unimpeded solar illumination of the cells while preventing moisture and atmospheric contaminants from attacking the solar cell material. Accelerated aging by operating the solar cells in an elevated temperature environment will be set up.
Undergraduate Role: The undergraduate will assist either a senior graduate student or post doc in setting up the tests, taking measurements and analyzing data.
Ultrawideband communication systems transmit information using very wideband signals. The minimum bandwidth is 500MHz but could be as large as 7.5GHz. The frequency band allocated by the FCC overlaps the frequency bands allocated to many other systems. The transmit power levels are extremely small so as to not interfere with these other systems. Interference from other systems is a significant challenge in the design of ultrawideband systems. This summer project involves simulation and possible implementation of interference mitigation techniques in ultrawideband communication systems. Depending on the experience of the student the project can involve evaluating performance of ultrawideband communication systems with interference mitigation via simulation or potentially could involve designing algorithms in either FPGA or DSP that mitigate interference.
Witricity: Cutting the Cord (2010)
Student: Tony Chen
Faculty Mentor: Anthony Grbic
This project involves hands-on experience in the emerging field of "witricity" (wireless electricity). Recently, the concept of transferring energy wirelessly through near-field coupling was proposed and experimentally verified by Soljacic and coworkers at MIT. In the experimental demonstration, power was efficiently transferred over a distance of 2 meters using two receiver and transmitter coils. The operating frequency of 9 MHz allowed the coils to remain within each other's near field even at distances of a few meters. Such a scheme could be used to efficiently charge mobile devices such as laptop computers, PDA's, digital cameras and cell phones wirelessly! The mobile devices would gradually charge throughout the course of the day, thereby removing the need for a power cord connection. Just imagine, not having to worry about plugging in your cell phone!
In this project, the student will assist Prof. Grbic and a graduate student in the development of wireless non-radiative power transfer (WNPT) systems. The student will first learn the fundamentals of WNPT and then perform circuit simulations of a system using Agilent Advanced Design System (ADS): an RF/microwave CAD package that is the industry leader in high frequency design. The second part of the project will involve designing, fabricating and testing innovative "witricity" prototype systems. The student will fabricate the printed-circuit prototypes in the microwave and millimeter wave fabrication laboratory using photolithographic methods. Testing will be performed using state-of-the-art microwave test equipment in the Microwave Metamaterials Laboratory. The end goal is to experimentally demonstrate the wireless charging of an electronic device.