|
|
|
University
of Pennsylvania
Center for Sensor Technologies
Sample of SUNSET Projects
Tentative List
The following descriptions give a sample of proposed
research projects. The main focus of the projects is sensor-related phenomena,
the generation of scientific knowledge, and its application to sensor
development. The criteria for selecting the projects are the educational
value, appropriateness of the project for undergraduate students,
and prospects for continuation after the summer. Also, projects
have an active involvement of graduate students in addition to faculty
so that graduate students can serve as role models for the SUNSET student.
In certain cases a technician will offer short workshops on the use of
specialized equipment. Over the course of the summer, this technician
will also be available for help and guidance for the students at no cost.
Once you have been admitted to the program, it is recommended
that you contact the faculty member for additional information on any
of these or other projects. Feel free to make arrangements with
the faculty member prior to starting the SUNSET program.
Sensors for Orthopaedic Biomechanics Applications
Professors Dawn Elliott and Robert Mauck of the Orthopaedic Surgery and Bioengineering Departments have a long track record of providing research projects to undergraduate students and integrating them in their overall research activities. The work has regularly resulted in papers with undergraduates as co-authors, highlighting the quality and level of involvement of the undergraduate students. The following projects will be conducted at the McKay Orthopaedic Research Lab, which consists of a multi-disciplinary group of faculty, graduate and undergraduate researchers. The undergraduate participants will attend weekly group meetings with the PI and graduate students and staff involved in specific projects. Additionally, monthly Orthopaedic Bioengineering seminars are scheduled throughout the summer, highlighting work being done internally within McKay, or via talks given by invited speakers from around the country and world. This creates a stimulating atmosphere for the undergraduate students, who have the opportunity to participate as full-fledged members of the group.
- Professor Dawn Elliott (Biomechanics of soft tissue; Dept. of Orthopedic Surgery and Bioengineering)
- delliott_at_mail.med.upenn.edu
- Research: disc biomechanics: progression of intervertebral disc generation and evaluation of disc treatments.
- Projects: see also Dr. Elliott's website: http://www.uphs.upenn.edu/orl/people/elliott/
- Sample Projects:
Quantifying tissue fiber alignment under tensile load (Prof. Dawn Elliott)
Orthopaedic soft tissues, including tendon, ligament, meniscus and disc, are composed of collagen fibers which have a distinct structural alignment. This structural alignment is critical for the tissue function, providing composite fiber-reinforced behavior that is anisotropic. During mechanical loading, these fibers rearrange under applied load. While current mathematical models incorporate initial fiber alignment, they do not account for rearrangement of fibers, because the alignment has not been measured. The objective of this project is to use quantitative polarized light microscopy in conjunction with established tensile mechanical testing protocols to quantify initial and changing fiber alignment distribution under tensile loading.
The student will select the appropriate imaging equipment and interface it with existing tensile testing equipment. The student will develop simultaneous tissue testing and imaging, and then will write a program to analyze the spatial and load-dependent fiber alignment. This project will give the student the opportunity to learn about and integrate concepts in mechanical testing, Matlab programming, and polarized light microscopy.
Nguyen AM, Sounok S, Johannessen W, Clark CC, Elliott DM, “Quantification of genipin crosslinking with free amino residues,” Summer Bioengineering Conference, Amelia Island, FL, 2006 (won Und, student paper prize)
Guerin HL, Baxter S, Nguyen A, Elliott DM, Villarraga ML, Kurtz SM, “Human intervertebral disc cartilagenous endplate biphasic mechanical properties”, Transactions of the Orthopaedic Research Society, 32:1135, 2007
The Cell as a Sensor: Spatial Resolution of Chondrocyte Response to Mechanical Signals for Cartilage Tissue Engineering (Prof. Robert Mauck)
Tissue engineering is the process of generating new tissues from combinations of cells, scaffolds, and specialized culture environments. One recent focus of cartilage tissue engineering has centered on functional tissue engineering (FTE) that utilizes bioreactors to recreate a physiologic loading environment and foster tissue growth. Most studies of cartilage FTE have focused on the use of a particular mechanical signal applied in isolation. The natural loading environment in joints is more complex, however, and changes with developmental stage. In order to create a realistic environment, a novel sliding contact bioreactor is under development to test the hypothesis that the long-term application of physiologic sliding contact will increase the compressive and tensile mechanical properties of chondrocyte-laden hydrogel constructs. To determine the mechanism by which cells adapt to their mechanical loading environment, this project will focus on the short-term response of chondrocytes to this novel loading environment. By coupling cell-level activity with models predicting local mechanical events, we will better appreciate the role of the cell as an in situ sensor and director of tissue growth and remodeling.
Students will be involved in the design and validation of the novel mechanical bioreactor systems. They will apply CAD techniques as well as computer interfacing with motion controllers and sensors. Students will also learn how to acquire and expand cells and grow them in monolayer and in 3D culture systems.
Nerurkar NL, Nguyen AM, Elliott DM, Mauck RL, “Annulus Fibrosus Tissue Engineering with Aligned Electrospun Nanofibrous Scaffolds,” Transactions of the Orthopaedic Research Society, 32:249, 2007.
Huang AH, Yeger-McKeever M, Yuan X, Mauck RL, "Direct Measurement of Tensile Properties of Chondrocyte and MSC-Laden Hydrogels," 2007, 53rd Annual Orthopaedic Research Society Meeting, San Diego, CA, 32:345.
Micro and Nano Electromechanical Sensors and Structures
A group of faculty members, including Professors H. Bau, A. Johnson, C, Kagan, J. Lukes, G. Piazza and J. Santiago, have an active research program dealing with nanoscopic materials, structures, and technologies for building nano-scale devices and sensors. This group also has a long tradition of providing undergraduate students with high-quality research projects, as demonstrated by the papers co-authored by undergraduates and track records of student participants continuing on to graduate school
- Professor Haim Bau , Dept. of Mechanical Engineering
The micro and nanofluidic laboratory at Penn focuses on the development of fully integrated, miniaturized laboratories (lab on chip) for disease diagnostics at the point of care, for drug screening, and for fundamental studies in biology. For example, one of our current NIH-sponsored projects focuses on rapid detection of the HIV virus in saliva samples. Additionally, the lab is developing pumps, stirrers, thermal cyclers, cell lysis chambers, nucleic acid purification and isolation reactors, and multi sensor arrays for the processing and detection of proteins (i.e., antigens and antibodies) and nucleic acids in body fluids. Our efforts are interdisciplinary in nature integrating concepts from fluid mechanics, electrostatics, magnetics, reaction kinetics, biology, and medicine. We have close collaborations with colleagues in biophysics, medicine, and material science.
Throughout the years, the lab has hosted a few undergraduate researchers who participated in variety of projects. For example, the two summer 2006 REU students developed a novel paraffin-based, magnetically-driven piston that doubles as a valve and a pump. This summer’s undergraduate researcher is developing a novel reactor for DNA amplification. All the reagents needed for the process are stored in the reactor’s chamber and released prior to use. On occasion, undergraduate researchers have co authored research papers. For example, Tom Murray assisted us in developing a novel process to fabricate a carbon nanopipe at the tip of pulled glass capillary. In recognition of his contribution, Tom was listed as a co-author of an archival publication (Kim et al., 2005). In the future, we will continue to host undergraduate researchers in our lab. We typically pair the undergraduate researcher with a graduate student or a postdoctoral fellow who acts as a mentor and provides close guidance.
Kim, B. M, Murray, T., and Bau, H. H., 2005, The Fabrication of Integrated Carbon Pipes with Sub Micron Diameters, Nanotechnology, 16, 1317-1320.
- cjohnson_at_physics.upenn.edu
- Research: Nanotechnology; molecular electronics, nano-sensors; nanotube electronics; Nanometer-scale transport properties
- Projects: see Dr. Johnsons's web site: http://www.lrsm.upenn.edu/~nanophys/
Sample Project
Biomolecule/Nanotube Hybrids for Chemical Sensor Applications (Prof. A.T. Charlie Johnson)
The group of Professor A.T. Charlie Johnson creates carbon nanotube devices functionalized with a variety of biomolecules for use as vapor and chemical sensors. Recently we have extended these approaches to other carbon nanostructures including graphene and reduced graphene oxide. For example, his group recently demonstrated that carbon nanotube field effect transistors functionalized with single-stranded DNA have many properties making them ideal for use in an electronic nose sensor system. Vapor responses of these sensors are controlled by the sequence of the single-stranded DNA, making it possible to generate a large number of sensors (hundreds or thousands) with different responses, as needed for an electronic nose system with performance rivaling biological olfaction. Additionally, nanotube devices functionalized with proteins show an electrical response to the binding of the complementary protein, but no response when the device is exposed to a negative control. We are working to demonstrate the use of such devices for the detection of biomarkers associated with cancer and other diseases. Students in the project will learn how to grow carbon nanotubes by catalytic chemical vapor deposition. They will fabricate nanotube devices using optical and/or electron-beam lithography. They will functionalize these devices and test their electrical responses in ambient and upon exposure to analytes. This project will thus expose participants to a number of important issues in nanoscale science, device physics, chemistry, and systems engineering.
Enrique Rojas , "Biomolecular Doping of Single-Walled Carbon Nanotubes by Thyroid Hormone", APS March Meeting, Montreal, Canada, (2005)
Enrique Rojas and Charlie Johnson, "Carbon Nanotube-Based Biodetection Of Triiodothyranine", National McNair Research Conference Abstracts, vol. October, (2003)
- Professor Cherie Kagan, Dept. of Electrical and Systems Eng; Dept. of Materials Science and Engineering.
- kagan_at_seas.upenn.edu
- Research: Application of molecular and nanoscale materials in transistors and memory devices, photovoltaic devices, and chemical and biological sensors (see also http://www.seas.upenn.edu/~kagan/)
Sample Project
Semiconducting organic thin film and nanowire transistors for chemical and biological sensing
Our research is focused on using organic thin film transistors or molecularly functionalized semiconductor NW transistors for electrical detection of biomolecular analytes. Flourescence measurements are commercially used to detect biomolecular species. Efforts to improve the sensitivitiy of optical methods are reaching the single molecule level in the research lab. Electrical detection offers a route to extend and complement optical methods of biomolecular sensing, avoiding the need to tag biomolecules with fluorescent labels and provides a more compact detection device. We used modulations in device current levels to probe the presence or interaction between biomolecular species. Molecular species provide lock-and-key recognition to allow discrimination between biomolecular species and selectivity difficult to achieve in sensing devices.
- Professor Jennifer Lukes; Dept. of Mechanical Engineering and Applied Science
Sample project:
Sensing Techniques for Phase Change Cooling of Electronics using MEM structures (Prof. Jennifer Lukes)
Removal of excess heat is a critical and escalating challenge for electronic systems such as high performance computers, power electronics, and radar systems. Cooling technologies based on liquid to vapor phase change (i.e. boiling) are among the most promising potential solutions to the electronics cooling problem due to the large latent heat absorption involved. Existing models to quantify the amount of cooling provided by phase change heat transfer are heavily empirical and rely on such parameters as bubble generation rate, bubble size, and bubble density. The goal of this summer project is to set up and perform sensing experiments under controlled conditions to isolate these parameters. Through this project the student will learn the basic principles of phase change heat transfer. He/she will also become familiar with basic microfabrication techniques and learn to build a prototype vapor chamber with microfabricated metal heaters. The student will then perform experiments on the vapor chamber at varying heat fluxes and use the preferred sensing technique(s) to detect bubble departure, bubble size, and bubble generation rate.
- Professor Gianluca Piazza: Dept. of Electrical and Systems Eng.
- piazza_at_seas.upenn.edu
- Research: piezoelectrical micro and nano systems for Wireless Communications, Biosensing and Medical Ultrasound applications
- Projects: See Dr. Piazza's website: http://www.seas.upenn.edu/~piazza/
Sample SUNFEST Projects:
a. AlN NEM Resonant Sensors (Prof. Gianluca Piazza)
The Micro and Nano Systems (PMANS) Laboratory under the direction of Professor Gianluca Piazza is developing Nano Electro Mechanical (NEM) resonators for chemical sensing applications. The goal is to obtain ultra-small devices capable of unprecedented mass detection in the order of zeptograms. The resonators are formed by a 100-200 nm thick AlN piezoelectric layer sandwiched by two metal electrodes. The key challenge for the realization of high frequency (10-40 GHz) nanoresonators consists in the definition of nanometer size features in the plane of the AlN film. The scope of the project is to reliably pattern nanometer size features of AlN piezoelectric film. Features ranging from 100 to 500 nm need to be defined either via focused ion beam, electron-beam lithography or nano-imprinting. As part of the overall research project, the student will design a set of experiments to characterize the capability and limitation of the aforementioned equipment. Statistical variation analysis will be conducted on the collected data to predict the equipment accuracy and then relate it to device performance. This project will allow the student to learn the principle of operation of state-of-the-art equipment for nanofabrication. He/she will become familiar with nanomechanical sensors and understand the principle of operation and ultimate limits of nanomechanical resonant devices.
b. Nano-Electromechanical Switches (Prof. Gianluca Piazza)
The PMANS Laboratory is also developing piezoelectrically-transduced NanoElectroMechanical (NEM) switches that will be employed for mechanical computing. Continuous advancement in the IC industry has scaled transistor dimensions in the nanoscale and by doing so has reduced the voltage supply values to 1 Volt or less. Operation of current transistors results in large power consumption in the off state. The main motivation of this project is to replace current transistors with nanomechanical devices to reduce power consumption. A key aspect in the realization of these NEM devices is their electrical characterization.
This project deals with the electrical characterization of NEM switches. The goal is to characterize the response of fast (1 GHz) switches by looking at frequency and time response of the device. The student will be involved with the design of the electronic setup that interfaces to the NEM switch. He/she will also develop an automated system for actuation and data collection. Finally, the student will analyze the experimental data and compare it to the theoretical prediction.The student will learn the principle of high frequency testing equipment and become familiar with the operation of NEM devices.
G. Piazza and A.P. Pisano, “Two-port stacked piezoelectric aluminum nitride contour-mode resonant MEMS”, Sensors and Actuators A-Physical, vol. A 136, pp. 638–645, 2007.
C. Zuo, N. Sinha, M. B. Pisani, C. R. Perez, R. Mahameed, G. Piazza, “Channel-Select RF MEMS Filters Based On Self-Coupled AlN Contour-Mode Piezoelectric Resonators”, to be presented at IEEE Ultrasonics Symposium.
- Professor Jorge J. Santiago-Aviles, Dept. of Electrical and Systems Eng.
- santiago_at_seas.upenn.edu
- Research: Micro and Meso-scale electromechanical systems; Materials for sensing applications; Nano-technology and sensors.
Sample SUNFEST Projects:
Electrochemical Formation of Supercapacitors
This project deals with the electrochemical formation of the active media for a super-capacitor. This active media is a composite of an electro-active polymer and carbon nano-fibers from electro-spinning and commercially obtained single wall carbon nano-tubes. The student should be interested in experimental materials science and / or physical chemistry. The student will be responsible for the electro-polymerization process, structural and electrical characterization of the resulting composite.
Modeling and simulation of Super-capacitors
This project deals with the mathematical modeling and simulation of a super-capacitor subjected to different heat inputs resulting from parasitic resistances and the electro-chemical redox process. The resulting thermal gradient couples with the elastic field (thermo-elasticity) deforming the device during operation. The student should be motivated by mathematical modeling and be skillful in math.
Displacement transduction using SERS and electro-spinning
This project deals with electro-spinning of nano-scale structures, mostly fibers loaded with various noble metals spheres with diameters no less than half the mean fiber diameter. The idea is to use plasmonics from the noble metal nano-particles to enhance the Raman signature (SERS, surface enhanced Raman scattering), and using the structures as mechanical transducers. The student must enjoy experimental work and be knowledgeable of chemistry at the general chemistry level.
Yu Wang, Linda Lamptey, Jorge Santiago-Aviles, "Large Negative Magnetoresistance in Electrospun Pregraphitic CarbonNanofibers Heat Treated Below 1000C", IEEE Tran. in Nanotechnology, vol. 2, (2005)
Miguel Perez Tolentino , Rogerio Furlan, Idalia Ramos, Jorge J. Santiago Aviles, "Fabrication of a meso combustor structure using electonics packaging ceramics (LTCC)", NCUR 2006, Asheville, NC, April 2006, p. 863
Jorge J. Santiago Aviles, Rogerio Furlan, Patricio Espinoza-Vallejos, and Miguel Perez-Tolentino, "Development of an LTCC based small combustor", Proceedings of CICMT 2006: IMAPS/ACerS 2nd International Conference and Exhibition on Ceramic Interconnect and Ceramic Microsystems Technologies, Denver Colorado, 2006, vol. , (2006), p. TP11.
Robotics and Control oriented projects
A large research effort is going on in the General Robotics Lab (GRASP), where a diverse group of faculty and students from computer, electrical, and mechanical engineering work together. The following projects are a sample of available research topics which will be done under supervision of faculty and will leverage the substantial research and training infrastructure in the GRASP Laboratory. This provides a natural setting for mentoring and inspiring students to embark on careers in engineering and in research.
- Professor Kostas Daniilidis; Dept of Computer and Information Science, Director Grasp Lab.
- kostas_at_cis.upenn.edu
- Research: - Tele-immersion, Robotics, Vision
- Projects: see Dr. Daniilidis web site: http://www.cis.upenn.edu/~kostas
Combination of visual and inertial sensors for assisting the visually impaired:
This project involves building and calibrating a system of a fish-eye camera and an Inertial Measurement Unit (IMU), rigidly attached to each other, but mounted
on a necklace or on the shoulder of a person. The project involves learning the function of an IMU, and calibrating the camera so that images obtained are rectified as if they
were always taken with the camera vertical axis parallel to the gravity vector.
- Professor Vijay Kumar: Dept. of Mechanical Engineering
Design of Optical Force Sensors for Micro-scale Assembly (Prof. Vijay Kumar)
This project involves the design of a system for automated assembly using a probe station. The goal is to automate the canonical problem of assembling a peg into a hole using probes at the end of a micro manipulator through pushing operations with feedback between the pushing operations from an optical microscope. Because off-the-shelf force sensors are expensive and cumbersome, we propose an optical sensor that consists of an elastic structure at the end of each probe. As in an atomic force microscope, we can calibrate the structure, and estimate the applied forces from the deformation of the structure. The ultimate goal is to integrate the force feedback into a control system for automated manipulation with quasi-static models for the assembly task. The student will become familiar with mechanical and optical setup and will learn the use of quasi-static models of mechanical systems with frictional contacts for planning micro-manipulation tasks.
D. Cappelleri, J. Fink, B. Mukundakrishnan, V. Kumar and J. Trinkle, Designing Open Loop Plans for Planar Micro Manipulation, Proceedings of the 2006 IEEE International Conference on Robotic and Automation (ICRA), Orlando, 2006.
- Professor Dan Lee; Dept. of Electrical and Systems Engineering
Sensors for Fast Automotive and Safe Robotic Vehicles (Prof. Dan Lee)
This project is part of the Ben Franklin Racing Team whose goal is to build fast, reliable, safe and autonomous vehicles that will revolutionize transportation systems in urban environments. We will leverage state-of-the-art advances in sensing, control theory, machine learning, automotive technology and artificial advantages to build robotic cars. We are currently integrating the sensors and drive-by-wire systems in the car. Research will involve constructing an array of laser, vision, and acoustic sensors for the car, as well as developing the underlying sensing, control, and navigational algorithms for the onboard computational cluster in the vehicle. This project will involve intense research and development in both the laboratory and testing in the field. Students with a background in mechanical design, electronics, programming, and project management will be able to participate in this exciting research project
S. Chitta, P. Vernaza, R. Geykhman, D. D. Lee, "Proprioceptive localization for a quadrupedal robot on known terrain ", IEEE In. Conf. Robotics and Automation, 2007, pp. 4582-4587.
For other projects, see Dr. Lee's web site: http://www.seas.upenn.edu/~ddlee
- Professor Dan Koditschek; Dept of Electrical and Systems Eng.
The recently established locomotion laboratories within the GRASP lab, directed by Professors Daniel Koditschek and Mark Yim, are investigating the role of smart, sensor-driven robotic structures on locomotion tasks.
Animals, unlike most current robotic systems, are able to change the physical properties of their limbs as well as their control parameters to help them adapt to changing conditions in their environment. The goal of the proposed project is to study (through experimentation) how leg stiffness and leg damping affect forward locomotion.
Data will be gathered by running experiments with our existing robotic locomotion platforms such as RHex, CkBot, and Edubot.
The student working on this project should be familiar with Matlab. Throughout this project, the student will learn about the capabilities and limitations of various sensing modalities, and how these affect the overall performance of robotic systems as they interact with the real world. The student will also get hands on experience running robotic experiments and learning about bio-inspired robotics. The student will work closely with either a graduate student or post doc throughout the project. Successful completion of this project will teach the student about subjects as diverse as: fabrication techniques, sensor fusion, optimization, experimental protocol, locomotion dynamics, and design methodology.
More information and movies can be found at:
http://kodlab.seas.upenn.edu/ResearchPage/index.php?leaf=155
- Professor Mark Yim, Dept. of Mechanial Engineering and Applied Mechanics
The following projects require mechanical, electrical, and/or computer
science skills.
Self-reconfigurable Modular Robots:
Robots that can rearrange their own topologies have been studied for
a few decades now (www.parc.com/modrobots). We are advancing the field
in three areas:
1) dynamic motions (think a monkey swinging through a jungle)
2) task specification (sure, you have a robot that can take anyshape, but
what shape should it take to solve a task?)
3) novel reconfiguring mechanisms (can we make tiny legos that rearrange themselves?)
Flying Robots:
We are developing a small indoor vehicle that uses a novel attitude
control mechanism. It is theoretically highly maneuverable yet also
intuitive to fly. Current work includes mechanical design of the
structure and test fixtures, simulation and characterization of
components, both in hover and in a wind tunnel.
Wireless Sensor Networks
- Professor Rahul Mangharam, Dept of Electrical and Systems Eng.
- rahulm_at_seas.upenn.edu
- Research area: Wireless Networking, Next generation medical devices,embedded systems, vehicular networks, parallel computing
- Dr. Mangharam's website: http://www.andrew.cmu.edu/~rahulm
Body Sensor Networks (Wireless)
In the next decade over 70 million Americans will be 65+ years. 8 out of 10 will have one chronic disease. The goal of this project is to design a flexible and wearable smart band-aid to monitor the patient's body vitals at all times. We are currently designing the hardware for heart (ECG) monitoring and would like you to help with the overall design and software. Each health-strip has a microcontroller, low-power radio, ECG and body activity sensors. The health-strip will monitor the patient and alert the doctor, ER and send all data to a remote database via the low-power transceiver. The system currently runs a real-time sensor operating system and will need to host a virtual machine so the device can be programmable remotely. This is a project in real-time embedded systems and wireless networks. If you have interests in embedded systems, operating systems and real-world hardware design, email Prof. Rahul Mangharam (rahulm@seas.upenn.edu). For more information see: http://www.seas.upenn.edu/~rahulm/docs/ZipCare.pdf
Vehicle-to-Vehicle (V2V) Wireless Networks
V2V multi-hop wireless communication holds the promise of making our driving experience safer, more efficient and more enjoyable. Every vehicle will soon be equipped with WiFi-like wireless interface and will participate on-road with other vehicles in a Peer-to-Peer network. By locally broadcasting safety messages, a disabled vehicle (i.e. airbag deployed) is able to significantly reduce the time to alert all oncoming vehicles. One goal of the inter-vehicle network is to alert and inform vehicles of on-road incidents so they may avert danger and avoid delays by selecting alternate routes. The focus of this project is to develop algorithms and network protocols for real-time traffic congestion probing. We eventually plan to deploy a large real network with several vehicles in Philadelphia. If you have interests in wireless networks, embedded systems and strong programming skills, email Prof. Rahul Mangharam (rahulm@seas.upenn.edu).For more information see: http://www.andrew.cmu.edu/~rahulm/Research/GrooveNet
- Embedded Virtual Machines (Wireless)
Embedded wireless networks have largely focused on open-loop sensing and monitoring. To address actuation in closed-loop wireless control systems there is a strong need to re-think the communication architectures and protocols for reliability, coordination and control. As the links, nodes and topology of wireless systems are inherently unreliable, such time-critical and safety-critical applications require programming abstractions where the tasks are assigned to the sensors, actuators and controllers as a single component rather than statically mapping a set of tasks to a specific physical node at design time. To this end, we introduce the Embedded Virtual Machine (EVM), a powerful and flexible programming abstraction where virtual components and their properties are maintained across node boundaries. In the context of process and discrete control, an EVM is the distributed runtime system that dynamically selects primary-backup sets of controllers to guarantee QoS given spatial and temporal constraints of the underlying wireless network. The EVM architecture defines explicit mechanisms for control, data and fault communication within the virtual component. EVM-based algorithms introduce new capabilities such as predictable outcomes and provably minimal graceful degradation during sensor/actuator failure, adaptation to mode changes and runtime optimization of resource consumption. Through the design of a natural gas process plant hardware-in-loop simulation we aim to demonstrate the preliminary capabilities of EVM-based wireless networks. . If you have interests in wireless networks, embedded systems and strong programming skills, email Prof. Rahul Mangharam For more information see: http://www.seas.upenn.edu/~rahulm/docs/evm.pdf
Deterministic Computation Models for Parallel Machines
By 2020, five years after you get your Ph.D., embedded processors will sport 4,096 cores, server CPUs will have 512 cores and desktop chips could use 128 cores. Universally, there will be a shift from existing sequential algorithms to parallel in domains spanning control systems, cryptography and process-intensive computation. But more importantly, there is a need to design and develop parallel algorithms and parallel computation architectures with deterministic operation. Given a large number of stochastic data-dependent input streams and a spatio-temporal variation of core utilization, the goal is to devise a class of algorithms and complementary hardware architectures for mission and safety-critical environments where unreliable software is not an option. If you have interests in parallel computing and strong programming skills, email Prof. Rahul Mangharam (rahulm@seas.upenn.edu). For more information see: http://www.seas.upenn.edu/~rahulm/docs/Real-TimeParallelComputing.pdf
Physcial Sensors for Medical and Energy Generation Applications
Professor Jay Zemel has extensive experience developing novel, highly sensitive flow and pressure sensors. The projects described below will familiarize students with a wide range of research activities including understanding the physics of operation as well as the fabrication technology, characterization, sensor conditioning and a variety of medical and other applications. The following description is a sampling of some of the projects that will be available to SUNFEST students. It highlights the multidisciplinary nature of the research.
- Professor Jay Zemel, Dept of Electrical and Systems Eng.
- zemel_at_seas.upenn.edu
- Also Prof. Babette Zemel, Children's Hospital of Philadelphia
- Research: Piezoelectric sensors for activity measurement, flow sensors.
Project 1: Feeding Monitor for Premature and High-Risk Neonates
For the past two years, a feeding monitor for premature and high risk infants has been under development in cooperation with Professor Barbara Medoff-Cooper (CHoP and School of Nursing), Dr. Joel Kaplan, (TracNatal, a start-up company). The need for such a feeding monitor has been well established for two decades. A novel, self-contained system is being developed that monitors simultaneously the suckling and breathing characteristics of these challenged infants. In order to prosper outside of a pediatric neonatal intensive care unit (NICU), an infant has to be able to receive nutrition by either breast or bottle feeding. Premature and at-risk infants frequently fo not have well developed suckling skills, a situation that can place them in mortal danger. By measuring and analyzing their feeding characteristics in a hospital setting, it may be possible to develop clinical procedures based on the measurements that significantly improve their survival and possible reduce the high costs of their treatment.
The system developed to date employs a commercial micro-electromechanical pressure sensor, a unique breathing sensor invented here and a compact microcontroller based system for electronically monitoring the sensor elements. These components have been assembled in a first generation apparatus that is no larger than a conventional nursing bottle.
The SUNFEST project would involve addressing a second generation system that is hermetically sealed to simplify its use, work with medical personnel to conduct initial studies on the information that can be obtained from the sucking and breathing records. The needed background for this project would be an electrical engineering, mechanical engineering, and/or a biomedical engineering major. This is a "hands-on" project requiring a strong experimental interest and, preferably, a course that uses microcontrollers.
-
Project 2: Energy-Neutral, Millifluidic, Combined Algae- Solar Cell System
The commercial development of algae based biofuels has been under consideration for many decades. A wide variety of bioreactors designs, growth methods (e.g., lighting arrangements, nutritional conditions and extraction methods), and species choices and modifications have been proposed. The goal of almost all of these proposed methods has been to provide optimal growth rates for algae considering the red-blue absorption arising from the properties of their chlorophyll ,. Recently, Gordon and Polle described a combination of bioreactor, solar photocell and high efficiency red LEDs that significantly enhance algal cell productivity. A review of this and other papers indicate that interest in a highly efficient combination of solar based micro-algal bioreactors and solar cells is of growing interest. The recent spike in fossil energy fuel prices, while short lived because of the current economic downturn, does not mitigate against the importance of addressing the need for what we refer to as energy neutral biofuel production.
We begin with a project where:
- the technologies for the major components already exist allowing systems to be designed and built to test various proposed hypotheses;
- insight into the mean time to failure (MTF) of the system exists;
- the energy costs for producing and recycling the system are readily determined;
- the only energy source for the system over its MTF is solar irradiance;
- and, there are reasonable prospects for a net electrical and chemical energy gain over and above the energy costs for manufacturing and eventually recycling the system.
If a system satisfying these five conditions could be developed, the minimum consequence would be an energy neutral system. An important issue is the economic cost of such an energy neutral system considering the requirement that future anthropogenic energy producing systems shall not add to the CO 2 load of the atmosphere. The proposed system is a fusion of both solar electric and biomass methods aimed at optimizing the use of incident solar irradiance. The approach adopted is to quantitatively account for the materials energy cost as well as the processing energy cost over the life of the system. It is a point that has been repeatedly addressed in recent years but is not always made a central concern. Detailed quantitative assessments of the energy cost for specific design choices will determine the environmental and economic viability. Included in the considerations will be the estimated time to failure, the gradual degradation of efficiency that inevitably arises and the energy cost of recycling the spent system.
The project is in its infancy. There are many, wide ranging issues that will need to be addressed. Students with broad engineering backgrounds and interests would be welcome. There are three professors involved with this topic at the present time; Professors David Graves (CBE), Jorge J. Santiago- Aviles (ESE) and Jay N. Zemel (ESE). If interested in this brief description, please contact Professor Zemel at zemel@ee.upenn.edu for further information and discussion.
Anatoly Gitelson, Hu Qiuang and Amos Richmond, "Photic Volume in Photobioreactors Supporting Ultrahigh Population Densities of the Photoautotroph Spirulina platensis", Applied And Environmental Microbiology, p. 1570–1573, 62, (1996)
Amos Richmond, "Biological Principles of Mass Cultivation", Chap. 8, Handbook of Microalgal Culture: Biotechnology and Applied Phycology, Blackwell Publishing Ltd, (2004)
Jeffrey M. Gordon and Juergen E. W. Polle, "Ultrahigh bioproductivity from algae",
Appl Microbiol Biotechnol., p. 969–975, 76, (2007)
An energy neural system is defined as one that can supply storable energy (e.g., a fuel product) entirely with solar energy but without energy input from fossil fuels. In this regard, solar, wind, tidal, geothermal, hydroelectric and conventional biomass methods could be viewed as energy neutral systems if suitable long term energy storage mechanisms can be developed
Optical and Vision Sensors
- Professor Jan Van der Spiegel, Dept. of Electrical and Systems Eng.
Various faculty members (Professors N. Engheta, E. Pugh and J. Van der Spiegel) and their post-docs and graduate students have a long tradition of research on biologically inspired sensors and systems. Over 65 undergraduate and 12 PhD students have been involved in the research program over the past several years. Several of these students were former SUNFEST fellows who are co-authors on papers and who have gone on to graduate school or are faculty members at others institutions. The goal of the overall research program is to study new approaches to vision sensors. The students will learn biological aspects of computational neuroscience as well as its engineering application.
Polarimetric Imaging System
One such project relates to polarization imaging. Polarization contains important information about the imaged environment, such as surface features, shapes, shading and roughness, which are ignored with traditional imaging systems. We are developing a focal plane imaging sensor capable of real-time extraction of polarization information including Stokes parameters [24-25] . The program has different aspects. The first one deals with the design and characterization of current-mode CMOS image sensors. Student will learn the operation of these new types of highly sensitive and high resolution sensors. They will be involved in the evaluation of the electrical and optical aspects of the seniors and will learn important principles of analog VLSI design, and device physics as it relates to image sensors. The second aspect is the fabrication of micro-polarizer arrays to be deposited on top of the custom designed image sensor. The student will be involved in the fabrication and characterization of the micro-polarizer arrays. He/she will also learn the principles of optics and polarization vision as well as automated measurement and analysis techniques. Students with interest in electric circuit design, materials, micro-fabrication technology as well as programming will able to participate in this ongoing research program and thus receive a well-rounded exposure to an exciting and rapidly growing field.
Image Sensor with Focal Plane Extraction of Polarimetric Information,” V. Gruev, J. Van der Spiegel and N. Engheta, Proc. IEEE Int. Symp. Circuits and Systems (ISCAS), pp. 213-216, May 2006
Viktor Gruev, Jan Van der Spiegel and Nader Engheta, “Real-time extraction of polarimetric information at the focal plane,” SPIE Defense and Security Symposium on "Polarization: Measurement, Analysis, and Remote Sensing VII" 17-22 April 2006, Orlando, Florida.
V. Gruev, K. Wu, J. Van der Spiegel and N. Engheta, "Fabrication of a Thin Film Micro Polarization Array", Proc. IEEE Int. Symp. Circuits and Systems (ISCAS) , (2006), p. 209
Viktor Gruev, Alessandro Ortu, Nathan Lazarus, Jan Van der Spiegel, and Nader Engheta., "Fabrication of Two Axial Thin Film Micro Polarization Array", Optics Express Letters , (2007), pp. 4994-5007
Physical Implementation of Computing
- Professor Andre' DeHon; Dept. of Electrical and Systems Engineering
- andre@seas.upenn.edu
- Research: Physical implementation of computation: physical substrates, programmable media, mapping, system abstractions and dynamic management, and problem capture
- Projects: See Dr. DeHon's website: http://www.seas.upenn.edu/~andre/
Parallel Floating Point Implementations -- With today's available silicon,
it is possible to connect numerous floating-point units in parallel to
potentially accelerate floating-point computations or preserve higher
precision/accuracy in floating-point computations. We envision a set
of related projects that build on this theme:
a. using parallelism to accelerate floating-point accumulation ---
we previously developed a strategy for using parallel hardware
to accelerate floating-point accumulation while providing the
same result as a sequential accumulation [2007];
reflecting on the initial design, there appears to be opportunities for
additional optimization which merit exploration and would make
a nice, self-contained project for a summer student.
b. using parallel floating-point accumulation hardware to perform
a "perfect" summation --- i.e. produce the correctly rounded
result as if the summation was performed with infinite precision
then rounded. Previous work shows how to implemented efficient,
sequential software solutions for this problem, but no work has explored
how to exploit parallelism and modern hardware for these cases or
assessed the potential benefits of hardware organizations tuned for this
purpose.
c. using parallel floating point units to distill a floating-point sum to
the minimum length multi-precision value --- i.e. represent the infinite
precision result with multiple limited-precision floating-point values;
again non-parallel software algorithms have been reported, but no work has
assessed the viability of aggressive parallel versions with appropriately
tuned modern hardware.
In previous work, we have engaged undergraduates in these kinds of task.
Stephanie Chen helped develop a parallel, saturated accumulator design
which was later published as [2005]. Jon Rameriz helped with
early versions of the ideas which became [2007]. In both
cases, we targeted an aggressive FPGA implementation to concretely
demonstrate the hardware designs. In all of the above cases, we have
identified the basic algorithms for the students to start with, but each
algorithm requires implementation, experimentation, and design space
exploration to determine if there is a benefit and to determine the
parameters and configurations which offer the greatest benefits. These
projects allow a student with basic math and computer programming
background to get an introduction to the concerns of parallel hardware,
computer arithmetic, and FPGA design, as well as experience with
engineering design space exploration.
Created by J. Van der Spiegel
Back to SUNFEST Homepage
Updated 01/22/09 |