ENABLING FEEDBACK FORCE CONTROL FOR COOPERATIVE TOWING ROBOTS

Advisors: Dr. Vijay Kumar and Jim Keller

ABSTRACT: Present research looks at the manipulation of a payload using multiple cooperative robots to drag it with cables to an assigned position within a given degree of accuracy. This project will increase the autonomy of simple tasks such as towing and will improve the efficiency of such systems. Moreover, these results will be applied towards improving cooperative technology and incorporated into the research of Jonathan Fink, et al [1] at the GRASP Lab. The project endeavors to accurately measure cable tensions between the payload and multiple cooperative robots in real time, transmitting this information to the user. We use cost-effective load cell sensors to measure tensile forces, outputting a change in frequency in proportion to the change in tension on the cell. Next, these signals are routed into the I/O pins of the BASIC Stamp 2 which wirelessly communicate this data to another BASIC Stamp 2 using an RF transmitter and receiver at 433.92 MHz. Our preliminary results show that we can wirelessly transmit information to a user, thus enabling crucial feedback to this system. This feedback would allow the robots to correct their position and velocity to maintain tension, allowing for less positional error in navigating the payload. View Paper | View Slides

GROOVENET3: A SCALABLE TRAFFIC SIMULATOR FOR CONGESTION PROBING AND PREDICTION

Advisor: Rahul Mangharam

ABSTRACT: Traffic congestion is a huge problem for developed countries such as the United States. Severe traffic congestion that can slow traffic up to several miles is common for everyday drivers in United States. Before the problem can be solved, government agencies and private agencies need reliable data to help understand traffic congestion. The goal of this research is to enable GrooveNet, a vehicle-to-vehicle simulation program, to record and analyze large amounts of historical traffic data, quickly and efficiently, and provide a playback capability on historical data. Methods used to develop this robust capability were proposed by Kanul, et al, at Carnegie Mellon University. Using historical traffic data, GrooveNet displays congestion over time and proposes a less congested route for a simulated vehicle. By gathering historical data, the upcoming version of GrooveNet will have a large knowledge repository that will ultimately give way to sophisticated machine learning and prediction algorithms that allow GrooveNet to navigate vehicles from origin to destination, safely and quickly, thus decreasing traffic congestion. View Paper

COMPACT ATTITUDE SENSOR SYSTEM USING SR-UKF

Advisor: Dr. Mark Yim

ABSTRACT: This paper describes the selection and integration of attitude sensors for a novel flying device. The overall goal of the experimental flyer project is to create a small, cheap, and lightweight device that is capable of controlled flight. This project focuses on the development of a state estimation system for the flyer that allows for precise and rapid attitude measurement. A sensor package consisting of a three-axis magnetometer, a three-axis accelerometer, and three one-axis rate gyroscopes is presented. The sensor measurements are filtered and fused together using a Square-Root Unscented Kalman Filter (SR-UKF) developed in MATLAB. The feasibility of porting the filter to a Microchip dsPIC30F microcontroller is discussed, and the performance of the filter is evaluated. View Paper | View Slides

EVALUATING AN INTERLEUKIN-1b INJECTION TO INDUCE DEGENERATION IN THE RAT LUMBAR INTERVERTEBRAL DISC

Advisors: Amy Orlansky, MS, and Dawn Elliot, PhD

ABSTRACT: Early signs of degenerate intervertebral discs include a decrease in the glycosaminoglycan (GAG) content within the nucleus pulposus and the annulus fibrosus, which has been linked to an alteration in the mechanics of the disc. However, the mechanisms for this reduction in GAG are currently unknown. An animal model that demonstrates disc degeneration would be useful not only for understanding the process of degeneration, but also for exploring methods of treatment or regeneration. Specifically, this study aimed to develop an in-vivo murine model resembling early intervertebral disc degeneration by introducing Interleukin-1β (IL-1) into the lumbar spine. IL-1 was injected into the nucleus pulposus of the lumbar intervertebral discs of rats, and these discs were evaluated using mechanical testing, biochemical assays, and histology at one and four weeks post-injection. A control group and a sham injection of PBS were used to evaluate the significance of the IL-1 injection. At one week, a reduction was seen in nucleus GAG content in the IL-1 discs. At four weeks, glycosaminoglycan restoration was seen in the IL-1β treated discs, suggesting the possibility of recovery of the disc. View Paper | View Slides

DYE-SENSITIZED ZNO FIBERS FROM ELECTROSPINNING AND PHOTOVOLTAIC CELLS

Advisor: Jorge Santiago-Aviles

ABSTRACT: Dye-Sensitized Solar Cells (DSC) have been studied because of the ability to transform solar into electrical energy at low cost. The system that DCS uses is a photo-electrochemical system based on a semiconductor sandwiched between a photosensitized anode and cathode, both immersed in an electrolyte. First it was important to develop a semiconductor for the DSC. Zinc oxide is a good semiconductor, and its synthesis is neither difficult nor expensive. Zinc oxide was produce using Poly(ethylene oxide), zinc acetate, water and anhydrous ether. Then we used the electrospinning technique with the solution to obtain fibers. These fibers were heated at 650 oC for 6 hours. For the characterization of the fibers we used Optical Microscope, Scanning Electron Microscope (SEM), and Raman Spectrometry. With the optical microscope we conclude that with the solution fibers could be made. By the shape of the fibers in the pictures taken with the SEM we could say that the fibers look like zinc oxide fibers. Finally with the Raman Spectrometry analyses we conclude that the fibers were compose by zinc oxide. View Paper | View Slides

NEONUR: A FEEDING DEVICE FOR PREMATURE NEONATAL NURSING

Advisor: Dr. Jay N. Zemel

ABSTRACT: Premature infants lack the ability to communicate verbally; therefore doctors who care for them need tools to measure their health, particularly their ability to nourish themselves adequately. These measurements need to be obtained in a variety of locations, such as intensive care nurseries and in homes. The bulky size and tethered nature of current devices measuring infant feeding make such monitoring too difficult. Nurses doing research at the Children’s Hospital of Philadelphia (CHOP) have requested the University of Pennsylvania School of Engineering to develop a device that will meet the needs of CHOP nursing researchers. During the 2007-2008 academic year, University of Pennsylvania Seniors Leslie Chen and Preeti Rajendran have worked on developing a prototype for a new system that has progressed to the troubleshooting phase. In particular, problems with the on board FLASH memory chip persist. This summer, functionality of data acquisition, data amplification, and Enhanced Universal Synchronous Asynchronous Receiver Transmitter (EUSART) communication between the microcontroller and a personal computer have been tested, modified and are now confirmed to be functioning properly. There is still a need for more testing before a prototype will be ready for further testing and optimization. View Paper | View Slides

GROWTH OF CARBON NANOTUBES VIA CHEMICAL VAPOR DEPOSITION

Advisors: Zhengtang Luo, A.T. Charlie Johnson

ABSTRACT: Carbon nanotubes (CNTs) are an unusual, tubular form of carbon, composed of a lattice of carbon hexagons rolled into a tube of nanometer-scale diameter. After the finding of CNTs in 1991 by Iijima on the cathode of an arc-discharge instrument, nanotechnology and nanoscience expanded tremendously. The subsequent discoveries of carbon nanotubes’ unique properties fueled research into all aspects of CNTs, from fundamental physics, to synthesis techniques, to technological applications. However, synthesis remains a challenge in almost every aspect, including control of length, diameter, and orientation. Also elusive is control of resulting nanotubes’ electric properties, which directly depend on variations in molecular structure. The other major challenge is production on a large scale, also currently far from accomplishable. Thus, we are concerned with optimizing methods to produce more consistent and higher yields. Specifically, we used chemical vapor deposition (CVD) with a focus on maximizing CNT length, alignment, and density of tubes per area. CNTs of lengths around 1mm have been achieved on silicon wafer chips, and two parameters – method of catalyst application, and catalyst concentration – have been optimized to achieve CNTs of this length. Additionally, moderate alignment of CNTs was achieved on ST-cut quartz chips and an apparent correlation between catalyst particle size and nanotube growth was found. View Paper | View Slides

DEVELOPMENT OF A MICRO PCR REACTOR FOR LAB-ON-A-CHIP DEVICES

Advisor: Haim H. Bau

ABSTRACT: The process of performing a Polymerase Chain Reaction (PCR) on a lab-on-a-chip (LOC) device is difficult to achieve. When treating a patient, a timely and accurate result is needed, but obtaining those results is challenging when using a conventional PCR machine. However, if the ability to detect molecules such as DNA during the PCR process happening inside the LOC device was accomplished, then the desired results would be obtained. One potential solution is the use of Real Time-Polymerase Chain Reaction (RT-PCR). Using SYBR Green as the principle dye for the RT-PCR process, this paper discusses the use of acrylic chips and RT-PCR tubes, exposed to different experimental parameters, being analyzed with the RT-PCR technique. According to our studies, better signals can be obtained from a RT-PCR melting curve with the use of FTA membranes than with the use of other membranes such as alumina and Porex® membranes. Good results were obtained from samples taken from chips that have experienced a PCR process with and without FTA membranes. Also, the acrylic plastic chip did not show signs of leakage or any kind of plastic damage while using normal PCR program settings. More importantly, RT-PCR tubes and chips were exposed to the different wash and incubation steps that would be performed on a LOC. Through the RT-PCR machine, good melting curves results were obtained from the vials but for the chips, further research is needed. View Paper | View Slides

Optimizing Legged Locomotion Using Tunable Leg Stiffness

Advisors: Dr. Dan Koditschek and Kevin Galloway

Abstract: Running efficiently and successfully over unstructured terrain is an important and necessary characteristic for modern legged robots. Research suggests that a necessary step toward achieving this goal is to design legged robots with variable leg stiffness capabilities which can be controlled by the robot autonomously. The specific problem addressed in this research is to develop a wireless (infrared) tunable leg stiffness module that will change leg stiffness when commanded by the robot. In this research, a legged robot with six C-shape legs called RHex is used. The development of the tunable leg stiffness module had two phases. The first phase focused on the design of an infrared communication module that would allow each leg to communicate with the robot body. The second phase involved the integration of a motor and rotary sensor to accurately control stiffness of the leg based commands from robot. In this paper, we discuss the development of this novel tunable robot leg and highlight its advantages and limitations for optimizing RHex-like legged locomotion platforms. View Paper | View Slides

TALKING TO ROBOTS: SPEECH-DIRECTED MOTION PLANNING

Advisor: Dr. George J. Pappas

ABSTRACT: In order to control a robot with spoken commands, concepts from speech recognition and motion planning must be integrated. This paper provides a solution to this problem, drawing on the Spoken Language Understanding Shell system developed by Schuler, and on recent work by Kress-Gazit et al. in temporal logic motion planning. Specifically, modifications and additions to the software are described, and excerpts provided, to demonstrate how the two systems are interfaced. Furthermore, successful performance of the new system in computer simulation is exhibited, allowing the user to direct a robot in complex scenarios with a minimal number of utterances. Finally, future work is discussed in multi-robot and environment-exploration applications. View Paper | View Slides

INVESTIGATION OF MSC DIFFERENTIATION ON ELECTROSPUN NANOFIBROUS SCAFFOLDS

Advisor: Robert L. Mauck and Nandan L. Nerurkar

ABSTRACT: Tissue engineering has been utilizing electrospun nanofibrous scaffolds as a platform for growing mesenchymal stem cells (MSCs). Electrospinning allows for the fabrication of scaffolds with a controllable degree of fiber alignment. Previous work has established that variation in the degree of fiber alignment directly effects the shape of the cells seeded onto these scaffolds. This study explored the effect of scaffold fiber alignment, and hence cell shape, on MSC differentiation. Two groups were observed in this study—aligned and non-aligned scaffolds. The scaffolds were seeded with MSCs and grown in a chemically defined culture medium for ten days. The effect of alignment was inspected using fluorescent microscopy to visualize cell morphology, biochemical assays to measure DNA content and glycosaminoglycan production, and real time PCR to determine gene expression. It was hypothesized that aligned scaffolds, in which cells are elongated in shape, would result in an up regulation of collagen type I, a marker of fibrous tissue differentiation, while non-aligned scaffolds would induce an up regulation of collagen type II, a cartilaginous marker. The data supported the second part of our hypothesis, that collagen type II would be up regulated in non-aligned scaffolds. However, collagen type I was not different between the two scaffold types. Because collagen type I is not unique to fibrous tissues, more specific markers are currently being investigated, including tenomodulin and lysyl oxidase. In conclusion, scaffold disorganization promoted chondrogenic differentiation of MSCs when compared to highly aligned fibers. These results support the contention that cell shape plays an important role in the differentiation of MSCs along fibro-cartilagenous phenotypes. View Paper