SUNFEST at Penn

Professor Jay Zemel, Electrical and Systems Engineering

There are many situations where it is desirable to have a compact, mobile system that measures a number of physiological parameters for time periods ranging from 5 minutes to a week. Professor J. N. Zemel's group has developed a small and adaptive microprocessor based measuring circuit and related software to measure a number of physiologically important variables associated with the growth and development of children. This research has been done in collaboration with colleagues at the Children's Hospital of Philadelphia, the Penn School of Nursing and the Penn Department of Pediatrics. One version of the device is used to monitor the feeding characteristics of neonates during a 5 minute feeding session (sucking pressure and breathing) and another is to monitor the physical activity of young children over a week-long period. The proposed project will focus on a wireless based sensor readout; improved breathing sensors for neonates; system for monitoring the physical activity of senior citizens; miniaturization and power efficiency of the circuitry.

The combination of working on a project that involves learning the underlying physical principles of the sensors, electronic design and fabrication along with programming of microcontroller based projects provide students with valuable research experience and skills for further graduate work.

Contact information:

• Professor Jay Zemel, zemel@seas.upenn.edu
• Also Professor Babette Zemel, Children's Hospital of Philadelphia, zemel@email.chop.edu
• Research: Piezoelectric sensors for activity measurement

Professor Cherie Kagan and Jan Van der Spiegel, Electrical and Systems Engineering

Organic electronics takes advantage of the low-cost, low-temperature processing methods of organic materials to fabricate large-area, flexible circuitry. The field of organic bio-electronics is emerging as exciting applications of organic devices and circuits for sensing biological processes. While most studies are focusing on chemical sensing, we are constructing organic circuits to sense electric fields in human tissue. Current technology uses needle-like electrodes that grossly under-samples neurons in tissue and that is limited to a handful of electrodes. Organic electronics allows for large-area, three-dimensionally conformable, high-density, and multiplexed, sensors and addressing circuitry. We are developing organic circuits to sense and stimulate tissue to understand the structure and computation of the brain network and to develop diagnostics and therapeutic technologies, impacting a wide-range of disorders in tissue. This is in collaboration with Prof. B. Litt's group of the Bio-Engineering Dept.

Students will be involved in the design, fabrication, and characterization of organic sensing and stimulating circuits. Students learn the principles of organic FETs, novel fabrication methods, software to design circuitry, and DC and AC electrical characterization methods. Students will also be involved in testing their circuits in animal models and as the technology advances observing the behavior of tissue.

Contact information:

• Cherie Kagan, kagan@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/)
• Jan Van der Spiegel, jan@seas.upenn.edu
• Research: Microelectronics, analog and digital VLSI, microsensors and solid-state imagers (http://www.seas.upenn.edu/~jan)

Professor Robert Mauck, Bioengineering

Adult stem cells are a promising cell source for regenerative medicine and tissue engineering applications as they can be induced to differentiate into a number of different cell types (including bone, cartilage, and adipose tissue). To create tissue engineered constructs, these cells are placed in a 3D biomaterial environment and exposed to a variety of soluble signals to induce differentiation. In addition to these soluble cues, mechanical features of the microenvironment influence how cells differentiate, and the time course and strength of this lineage commitment. Our group generates advanced biomaterials for specific application to articular cartilage, knee meniscus, and intervertebral disc tissue engineering. Further, we develop custom mechanical bioreactor systems to stimulate the growth of these engineered constructs in the lab. Ongoing studies are exploring how adult stem cells carry out their role as "sensors" of the local mechanical environment, and seek to identify the molecular, mechanical, and signal transduction pathways that regulate the differentiation process.

Working with senior graduate students and post-doctoral fellows, students will be involved in the design and fabrication of materials for tissue engineering, and the operation of custom bioreactor systems to mechanically perturb tissue engineered constructs and evaluate stem cell response to these perturbations. The students will learn how to carry out basic tissue culture protocols, how to acquire and analyze mechanical test data, how to perform gene expression and histological analysis, and how to interpret their results in the context of cell differentiation pathways.

Contact information:

• Robert Mauck, lemauck@mail.med.upenn.edu
• Research: Tissue Engineering, Stem Cell Biology, Mechanobiology, and Mechanotransduction, Biomaterials, Bioreactors
• See also Dr. Mauck's website.

Professor A. T. (Charlle) Johnson, Physics and Electrical and Systems Engineering

Integration of modern nanoelectronic technology with the potent molecular machines of living organisms offers a pathway to advanced modalities for chemical sensing, high throughput screening of ligand binding, and other applications. The group of Professor A.T. Charlie Johnson is focusing on experimental measurements and advanced computer simulations of electronic devices based on nanocarbon species (carbon nanotubes, exfoliated graphene, and large-area graphene) that are functionalized with a variety of biomolecules for use as vapor and chemical sensors. For example, his group recently demonstrated that carbon nanotube field effect transistors functionalized with actual olfactory receptor proteins from mice (mORs) show vapor responses that vary with the identity of the mOR and agree with mOR responses measured in a biological surrogate. We would like to extend this approach to other members of the G-protein coupled receptor (GPCR) protein family, of which ORs are but one example. GPCRs play a critical role in the activation of cellular signal transduction pathways and are also the target of approximately 30% of all modern medicinal drugs.

Students involved in experimental projects will be exposed to modern methods of carbon materials synthesis, nanofabrication, and biomolecular functionalization. They will test the electrical responses of their devices in ambient and upon exposure to analytes. Students will learn all-atom molecular dynamics techniques implemented on advanced computational resources. The goal is to develop a quantitative, computational basis for understanding the properties of the hybrid nanosystems studied by the group.

Contact information:

• Charles Johnson, cjohnson@physics.upenn.edu
• Research: Nanotechnology; molecular electronics, nano-sensors; nanotube electronics; Nanometer-scale transport properties
• See Dr. Johnsons's website.

Professor Haim Bau, Mechanical Engineering and Applied Mechanics

The micro and nano fluidics laboratory at Penn studies fluid flow under the action of pressure, electric, and magnetic fields; develops tools for in-situ imaging of processes that take place in liquid media and develops fully integrated, miniaturized laboratories (lab on chip) for disease diagnostics at the point of care, for drug screening, and for fundamental studies in biology. Our efforts are interdisciplinary in nature integrating concepts from fluid mechanics, electrostatics, reaction kinetics, biology, and medicine. We have close collaborations with colleagues in biophysics, medicine, and material science.

Projects that REU students will be working on involve developing an electrochemical sensor for measuring the temperature of our nanoaquarium that consists of a 100nm thick fluid layer confined between two 50nm thick silicon nitride membranes. To determine the fluid's temperature, electrodes are patterned inside the thin flow cell to measure the liquid's electrical conductivity. The liquid's temperature is inferred from its conductivity. A related project is the development of a "thermal battery" that can sustain an exothermic electrical reaction for about an hour to maintain an adjacent reaction chamber at 65 oC. The elevated temperature is used to facilitate a Loop-mediated Isothermal Amplification Reaction (LAMP). The LAMP process is used to amplify nucleic acids associated with various pathogens such as malaria and HIV. Undergraduate researchers will be paired with a graduate student or a postdoctoral fellow who acts as a daily mentor and provides close guidance.

The projects address both basic science and major socio-economical issues. For example, our nanoaquarium allows imaging of processes taking place in liquid media in real time and with the high resolution of the electron microscope. We anticipate that the device would allow one to gain a better understanding of the formation of colloidal crystals for photonic and electronic applications and develop new processes. Our diagnostic devices have the potential of improving world health, in particular in developing nations. We educate students at cutting edge, interdisciplinary technologies and prepare them for research careers in an emerging field of nano-fluidics.

Contact information:

• Haim Bau, bau@seas.upenn.edu
• Research areas: micro and nano fluidics; active control of flow pattern and molecular motors.
• See Dr. Bau's website.

Professor Jorge Santiago-Aviles, Electrical and Systems Engineering

Bandgap engineering of nanoscopic materials is a promising technique for electro-optics applications. This project involves the generation of nanoscopic wide band gap semiconducting nanofibers from organometallic precursors and electrospinning . The materials will involve the ternary system Al - Ga - N (GaN-AlN). The project will engage the student in the synthesis and characterization of the liquid precursor, the deposition and characterization of the nanoscopic fibers in the green, and the processing of the fibers to obtain the polycrystalline aggregate. Optical, electrical and mechanical properties will be measured as to gauge the materials use as sensors. Process modeling and simulation will be carried out with colleagues at the Univ. of Puerto Rico. The student will learn all facets of electronic materials and device research, from synthesis to device implementation. The student will also be exposed to an interdisciplinary and multidisciplinary environment, working with chemists, physicists and engineers, as well as interacting with a diverse group of researchers from different ethnic backgrounds and universities.

Contact information:

• Jorge Santiago-Aviles, santiago@seas.upenn.edu
• Research: Micro and Meso-scale electromechanical systems; Materials for sensing applications; Nano-technology and sensors.

Professor Jorge Santiago-Aviles, Electrical and Systems Engineering

Energy storage schemes are highly thought after now that the nation is moving toward less reliance in fossil fuels. Supercapacitors alone and in conjunction with high performance electrical batteries are seen as reasonable scheme. This project involves the generation of Faradaic (using electro-active polymers, EAP) and DLC (double layer capacitance) electrodes and their combination into asymmetrical supercapacitors. This project will require that the student learn some fundamentals of electro-chemistry as to be able to electro-polymerize the EAP and characterize the supercapacitors using electrochemistry techniques of such as cyclic voltametry and impedance spectroscopy. The student will be involved in process and device modeling and simulations. The student will be exposed to a total interdisciplinary and multidisciplinary environment, working with chemist, physicist and engineers, as well as interacting with a diverse group involving researchers from different ethnic backgrounds and universities. 

Contact information:

• Jorge Santiago-Aviles, santiago@seas.upenn.edu
• Research: Micro and Meso-scale electromechanical systems; Materials for sensing applications; Nano-technology and sensors.

Professors Nader Engheta and Jan Van der Spiegel, Electrical and Systems Engineering

The goal of the overall research program is to study new approaches to vision sensors . One such project relates to polarization imaging. Polarization is becoming of increased interest in biological imaging and remote sensing. In contrast to intensity and spectral information, polarization provides information about surface orientation and roughness, and is only weakly dependent on the material parameters and overall scattering cross section of the objects in an image . However, measuring the polarization properties of a scene remains challenging due to the lack of on-chip high-quality microgrid polarizers as well as the lack of advanced microgrid based algorithms that can correct for artifacts associated with the specific nature of polarization imaging. Our research aims to develop a microgrid polarimetric imaging system that consists of nano-fabricated wire-grid polarizers integrated on a custom imager . REU students will be involved in optical imager design and testing, optimization of polarizer design and fabrication as well bio-inspired algorithms. The students will learn biological aspects of computational neuroscience, the physics of polarization as well as its engineering applications.

Contact Information:

• Nader Engheta, engheta@seas.upenn.edu
• Jan Van der Spiegel, jan@seas.upenn.edu
• Research: Analog VLSI and Smart Vision CMOS Sensors; polarization image sensors.
• See Dr. Engheta's website.
• See Dr. Van der Spiegel's website.

Professor Rahul Mangharam, Electrical and Systems Engineering and Computer and Information Science

According to the 2009 Urban Mobility Report, delays due to traffic congestion cost the nation $87 Billion in the form of lost hours and wasted fuel. Building highways is an expensive way to increase traffic capacity.

The goal for the project is to study real-time, dynamic routing algorithms for nodes in sparse graphs, such as street maps. The algorithm will consider current and known future conditions, and will also be able to react to unplanned incidents that occur.

The aim is to investigate the construction, instrumentation and scheduling of time-bounded and anytime algorithms on multi-core architectures such as graphics processing units (GPUs). Most algorithms are run-to-completion and provide one answer upon completion and no answer if interrupted before completion. On the other hand, anytime algorithms have a monotonically increasing utility with the length of execution time. Such imprecise and approximate computing has wide application in prediction algorithms in the domains of vehicle traffic congestion, stock price prediction and weather prediction. Our investigation focuses on time-bounded anytime algorithms on GPUs for real-time vehicle traffic congestion prediction and route assignment. To explore this, we have designed AutoMatrix, a traffic congestion simulation platform on the Nvidia CUDA-enabled GPU. AutoMatrix is capable of simulating over 16 million vehicles on any US street map. This research has the potential to extend real-time traffic scheduling on massively parallel GPU architectures to attack a variety of data-driven, interactive and dynamical algorithms with timely operation.

The project was initiated and developed by a series of undergraduate research students. For example, Danny Lustig (ESE'09), helped with the initial design and is now pursuing a successful Ph.D. at Princeton University. Students involved in this project will learn sensor programming, parallel computing, algorithm design for massively parallel programs and traffic modeling. The project exposes students to various sensors, the future of computer architecture, new methods of programming and traffic engineering for sustainable cities of the future.

Contact information:

• Rahul Mangharam, rahulm@seas.upenn.edu
• Research area: Wireless Networking, Next generation medical devices,embedded systems, vehicular networks, parallel computing
• See Dr. Mangharam's website.

Professor Rahul Mangharam, Electrical and Systems Engineering and Computer and Information Science

Designing bug-free medical device software is difficult, especially in complex implantable devices that may be used in unanticipated contexts. There is currently no formal methodology or open experimental platform to validate the correct operation of medical device software. To this effect, a real-time Virtual Heart Model (VHM) has been developed to model the electrophysiological operation of the functioning and malfunctioning heart. We demonstrated that the VHM is capable of generating clinically-relevant response to intrinsic (i.e. premature stimuli) and external (i.e. artificial pacemaker) signals for a variety of common arrhythmias. The VHM has also been implemented on a hardware platform for closed-loop experimentation with existing and virtual medical devices. The VHM allows for exploratory electrophysiology studies for physicians to evaluate their diagnosis and determine the appropriate device therapy.

Students involved in this project will learn the details of medical sensor and device design from concept to validation and certification with a firm grounding in the software design for safety-critical systems. Students will learn the importance of timing analysis of software and software verification techniques to ensure medical devices are safe within the human body. Allison Connolly, a former SUNFEST09 REU undergraduate from Johns Hopkins University, was part of the initial development of the project. She is the co-author of five publications and is proceeding to University of Minnesota for her Ph.D. in Biomedical Engineering.

Contact information:

• Rahul Mangharam, rahulm@seas.upenn.edu
• Research area: Wireless Networking, Next generation medical devices,embedded systems, vehicular networks, parallel computing
• See Dr. Mangharam's website.

Professor Daniel Koditschek, Electrical and Systems Engineering

RHex was the first legged machine to run autonomously in general outdoor unstructured terrain and in the decade since its invention, there have been a variety of incremental design iterations including a version for educational use, all promoting further varied experiments. The growing experience with these related designs and their performance make the robot a natural platform for a study of one of the major unresolved questions in robotics: how to couple the design of form and function. The modular robotics literature has made some progress toward unified representations for specially designed systems of component modules. A small literature in evolutionary and developmental robotics has addressed this central problem in the context of examining how to bring adaptive pressure to bear on representations of more general designs that combine form with function. However these earlier attempts have relied purely on simulation or quasi-static settings that miss the crucial element of work (energy exchange with the environment) that makes a robot's body so hard to engineer.

This project aims to use the existing variant RHex design files, simulation engines, and empirical tests as a database for building a "genome," a unified representation that links the CAD and algorithmic specifications at least up to the level of an automated assembly plan as could be subjected to human mediated rational design methods. Students will be involved in learning how to use and understand the design representations in the various databases and CAD systems that cumulatively represent the RHex designs, and then extracting and pulling together a linked, unified representation with accompanying database of working design instances. Continuations of the project will involve the use of the database to advance hypotheses about superior design points and attempt to test these empirically on variant robot components.

Contact information:

• Daniel Koditschek, kod@seas.upenn.edu
• Research: Robotics
• See Dr. Koditschek's website.

Professor Camillo Jose Taylor, Computer and Information Science

The goal of this work is to explore new ways for humans to operate advanced humanoid robotic systems. More specifically the aim is to develop a system that will allow a human user to virtually inhabit our newly acquired PR2 humanoid robot from Willow Garage. This system is sufficiently anthropomorphic to allow us to consider mapping the motions of a human operator directly onto the motions of the head, base and arms of the robot. The concept is to outfit the operator with a virtual reality headset, monitor his/her movements with a Vicon motion capture system and then map those motions onto the robot while relaying the video feeds from the robots head camera back to the head mounted display to create an immersive teleoperation experience. The project will involve the design and implementation of code that maps the human motion onto the robots.

Over the course of the summer the students will learn principles of robotics, various sensors and its control. They will learn Python, one of the core languages for programming our robot system. Student will also become familiar with the operating system, ROS, which runs the robots. In addition they will become familiar with the Vicon system used to capture human motion.

Contact information:

• Camillo Jose Taylor, cjtaylor@cis.upenn.edu
• Research: Computer Vision and Robotics
• See Dr. Taylor's website.

Professor Daniel Lee, Electrical and Systems Engineering and Computer and Information Science

The project will explore the use of laser and inertial sensors to simultaneously localize and map an unknown environment using an unmanned mobile sensor platform. The student will work on both simulated and hardware systems that incorporate LIDAR scanners, MEMS gyroscopes and accelerometers to estimate 6 degree of freedom position and orientation. The estimated pose variables will then be used to control a pan-tilt head containing a laser pointer to highlight an object of interest. The student will learn about probabilistic representations, homogeneous transforms, as well as gain experience with experimental hardware and computational platforms.

Professor Lee's group has a long tradition of undergraduate student involvement. He leads the University of Pennsylvania's soccer legged robot team, the UPennalizers in the annual Robocup competition, and the Darpa Urban challenge. His group has been selected recently by the U.S. Army jointly with Australia's Defense Science and Technology Organization as one of the 6 finalists to participate in the inaugural Multi Autonomous Ground-Robotic International Challenge ( MAGIC ). These projects involve many undergraduate students. Several of these have gone on the graduate school at Penn and elsewhere.

Contact information:

• Daniel Lee, ddlee@seas.upenn.edu
• Research: Machine Learning and Robotics
• See Dr. Lee's website.

How should we organize multi-billion transistor programmable chips? How do we best use today's programmable components to design even better next generation components? Modern computations architectures are complex and multidimensional. As engineers, we would like to thoroughly explore this space, but it is too large to approach exhaustively. Fortunately, advances in statistical modeling and inference suggest powerful mathematical techniques to more efficiently explore these large spaces. In this effort, the student will learn about statistical sampling, numerical methods, and FPGA architecture to develop next generation computer-aided design tools.

Inspirational prior work includes:

• Lee and Brooks. Spatial Sampling and Regression Strategies. IEEE Micro, Volume 27, number 3, pp. 74--93
• Lin, Wawrzynek, and El Gamal. Exploring FPGA Routing Architecture Stochastically. IEEE Transactions on Computer-Aided Design, Volume 29, Number 10, pp. 1509--1522

Research group home page:

• IC Research: http://ic.ese.upenn.edu/
• Led by: http://www.seas.upenn.edu/~andre/

Contact Information:

• Andre DeHon, andre@seas.upenn.edu
• Laboratory website: http://www.seas.upenn.edu/~andre/

Today's computers are brittle and insecure. This is clear from the steady stream of security breaches and the patch-of-the-week grind we find ourselves in as we attempt to prevent further compromises. We argue that the current state of affairs is an unsurprising result of the way processors and operating systems have been traditionally designed. Exploiting today's abundant hardware resources, we believe it is possible to change the game and build a new class of systems that are inherently less vulnerable. Particularly, by adding rich metadata to the hardware representation for primitive data, we can improve the security and safety of computations. This can remove vulnerability to common security breaches (e.g. buffer overflow, stack bashing) and guarantee common errors (misusing an integer as a pointer) are cleanly caught and reported. This ongoingproject involves the development and optimization of a novel processor, prototyping on an FPGA platform, the development of software to exploit this processor, formal definitions of intended behavior and properties, and verification.

Background papers for effort:

• PLOS paper: http://www.crash-safe.org/node/9
• TIARA: http://www.dtic.mil/srch/doc?collection=t3&id=ADA511350

Research Group: http://www.crash-safe.org/

Hardware effort led by: http://www.seas.upenn.edu/~andre/

Contact Information:

• Andre DeHon, andre@seas.upenn.edu
• Laboratory website: http://www.seas.upenn.edu/~andre/