SUNFEST at Penn

For Prospective Students:

  Apply Online
  Faculty and Areas of Research
  Sample Projects
  Program Alumni and Past Projects
  SUNFEST in the News
 SUNFEST Home

For Accepted Students:

  Required Workshops
  Calendar
  Housing - Intern. House Phila
  Organizational Matters
  Current Students & Projects
  Student FAQs
  Student Resources
  Library Resources

Contact SUNFEST:
sunfest@seas.upenn.edu

Mailing Address:
SUNFEST Program
203 Moore Building
University of Pennsylvania
200 S. 33rd Street
Philadelphia, PA 19104-6314

SUNFEST at Penn

2017 Sample Projects

Descriptions of sample projects are given below. Use these to select which project you would like to work on. All the projects are related to the general area of sensor technologies, which acts as a common, intellectual focus.

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 SUNFEST program.

Self-Cleaning Vapor Sensors Based on Nanoporous TiO2 Bragg Reflector (Prof. Daeyeon Lee, CBE) - daeyeon@seas.upenn.edu

Website: http://www.seas.upenn.edu/~leegroup/index.html

The goal of this project is to provide research opportunities to understand the principles of gas/vapor sensors based on nanoporous Bragg reflectors, which are colorimetric sensors that can detect the presence and concentration of various gaseous analytes.  This work is inspired by nanoporous TiO2 Bragg reflectors that our group recently developed using blocking layer-assisted spin coating [15].  The work involves fabricating multilayers of nanoporous TiO2 domains by sequentially spin coating TiO2 nanoparticles of different shapes and sizes.  To prevent the cross-contamination of different layers, we will use polystyrene sacrificial blocking layer. Capillary condensation in the presence of an analyte from the vapor phase will induce change in the refractive index of TiO2 layers, leading to shift in the Bragg reflection and thus color change of the Bragg reflector. We have already demonstrated the changes in the color of the Bragg reflector in the presence of water vapor (Figure below). Moreover, using the catalytic activity of TiO2, self-cleaning properties of these Bragg reflector sensors will be tested.
The student(s) will develop a new Bragg reflector and their response to different analytes will be tested.  The student will learn both the theoretical and experimental aspects of photonic materials from this project.  Prof. Lee has a strong history of mentoring undergraduate students (> 40 students since 2004) and has published over 10 papers with undergraduate researchers.

  • [15]     Y.-R. Huang, J. T. Park, J. H. Prosser, J. H. Kim, and D. Lee, “Multifunctional All-TiO2 Bragg Stacks Based on Blocking Layer-Assisted Spin Coating,” J. Mater. Chem. C, no. 2, pp. 3260–3269, 2014.
Biosensors based on 2-D Materials (A.T. Charlie Johnson, Physics & Astronomy, and EE) - cjohnson@physics.upenn.edu

Website: http://nanophys.seas.upenn.edu/research.html
Nanomaterials such as carbon nanotubes and graphene have great promise for biosensing applications due to their environmentally sensitive electronic properties and all-surface geometries that result in enhanced sensitivity to target binding events. Nanomaterial-protein hybrid devices have been used to detect biomarker proteins of cancer [16], Lyme disease [17] and other small molecule targets [18] at femto-molar concentrations. It has been suggested that two-dimensional materials that are intrinsic semiconductors with an energy band gap (e.g. MoS2) may exhibit better performance than graphene for this application [19]. This project is focused on understanding the operation of nano-biosensors and developing scalable fabrication methods to produce them. Students will be involved in chemical vapor deposition of two-dimensional materials (graphene, MoS2, BN, etc.), analysis of materials by Atomic Force Microscopy & Transmission Electron Microscopy micro fabrication and electrical measurements to assess device properties including sensor responses to target analytes. Prof. Johnson has a long tradition of working with undergraduate students, many of which are co-authors and have gone on to graduate school.

  • [16]     M. B. Lerner, J. D'Souza, T. Paina, J. Dailey*, B. R. Goldsmith, M. K. Robinson, and C. A. T. Johnson, “Hybrids of a Genetically Engineered Antibody and a Carbon Nanotube Transistor for Detection of Prostate Cancer Biomarkers,” Nano, no. 6, pp. 5143 – 5149, 2012.
  • [17]     M. B. Lerner, J. Dailey*, B. R. Goldsmith, D. Brisson, and A. T. C. Johnson, “Detecting Lyme Disease Using Antibody-Functionalized Single-Walled Carbon Nanotube Transistors,” Biosens. Bioelectron., vol. 45, pp. 163–167, 2013.
  • [18]     M. B. Lerner, F. Matsunaga, G. H. Han, S. J. Hong, J. Xi, A. Crook*, J. M. Perez-Aguilar, Y. W. Park, J. G. Saven, R. Liu, and A. T. C. Johnson, “Scalable Production of Highly Sensitive Nanosensors Based on Graphene Functionalized with a Designed G Protein-Coupled Receptor,” Nano Lett., vol. 14, no. 5, pp. 2709–2714, May 2014.
  • [19]     N. J. Kybert, C. H. Naylor, C. Schneier, J. Xi, G. Romero*, J. G. Saven, R. Liu, and A. T. C. Johnson, “Scalable production of molybdenum disulfide-based biosensors,” Nano ASAP, 2016.
Deploying Environmental Sampling Systems from Autonomous Air Vehicles. (Prof. CJ Taylor, CIS) - cjtaylor@seas.upenn.edu

Website: http://www.cis.upenn.edu/~cjtaylor/RESEARCH/research.html

We have been working on the design of insect traps and other data collection systems that can be attached to and deployed by autonomous aerial vehicles. Over the course of the next year we would like to move towards the challenging task of developing software that would guide an autonomous aerial vehicle to automatically deploy and collect these pods opening the way to more automated monitoring of fields and extended areas.

Nanocrystal-Based Plasmonic Nanoantenna Arrays as Soil pH Sensors (Prof. Cherie Kagan, EE) kagan@seas.upenn.edu

Website:  http://kagan.seas.upenn.edu/


The goal of this project is to create non-toxic, large-area, colorimetric, in-field sensors of soil pH for aerial and ground-based drone monitoring. Soil pH is critical to agricultural plant growth. It affects the concentration of dissolved metal ions, some of which are toxic to plants and others that are needed nutrients. This project builds on our recent demonstrations of the large-area fabrication of plasmonic nanoantenna arrays using nanoimprint lithography of colloidal nanocrystals inks followed by ligand exchange [20], [21]  and of their use in combination with hydrogels to serve as colorimetric humidity sensors. Here, we will combine pH sensitive hydrogels with our nanocrystal-based plasmonic nano-antenna arrays to form colorimetric sensors whose wavelength of operation will be tuned by selecting the dimensions of the nano-antennas to transparency windows for water and plant lignin and cellulose. We will measure the magnitude and kinetics of the shift in the frequency of the optical response from the antenna array as the dielectric constant of the hydrogel changes in response to different pH solutions and ultimately to different soil pH.
The student(s) will learn about nanoscale materials and device fabrication and their integration and characterization as optical sensors; student will become aware of the importance of technology in farming to address the challenge of food security. The Kagan group has welcomed well over 20 undergraduates during the past ten years and provided exciting research experiences to students with various backgrounds including EE, ME and Material Science. Several of these have gone on to graduate school.

  • [20]     C. R. Fafarman, A. T.; Hong, S.-H.; Caglayan, H.; Ye, X.; Diroll, B. T.; Paik, T.; Engheta, N.; Murray, C. B.; Kagan, “Chemically Tailored Dielectric-to-Metal Transition for the Design of Metamaterials from Nanoimprinted Colloidal Nanocrystals,” Nano Lett, vol. 13, pp. 350–357, 2013.
  • [21]       C. R. Chen, W.; Tymchenko, M.; Gopalan, P.; Ye, X.; Wu, Y.; Zhang, M.; Murray, C. B.; Alu, A.; Kagan, “Large-Area Nanoimprinted Colloidal Au Nanocrystal-Based Nanoantennas for Ultrathin Polarizing Plasmonic Metasurfaces,” Nano Lett, vol. 15, pp. 5254–5260, 2015.
Designing Temperature-Responsive Nanomaterials for Sensing Local Environment (Jennifer Lukes, Mech. Eng.) - jrlukes@seas.upenn.edu

Website: http://www.seas.upenn.edu/~jrlukes/

The idea of this project is to use computer simulations to design new temperature-responsive materials that sense the local environment and adapt their thermal  properties accordingly. Such materials will be beneficial in a variety of thermal management applications, including tailorable thermal insulations and self-cooling to prevent overheating in electronics and batteries. In this project, the student will analyze how the structure and composition of a nanocomposite material can be tuned to achieve a desired temperature-dependent thermal conductivity. This will be accomplished by making an initial educated guess as to what material structure and composition will produce the desired thermal conductivity characteristics, performing theoretical modeling to calculate the thermal conductivity of this material, and iterating to find the optimal structure and composition. This project is ideal for students who are more theoretically oriented. The student will learn about conduction heat transfer in materials and how to perform molecular dynamics simulations [22], [23], [24]. Prof. Lukes has previously mentored several undergraduate REU students on projects related to thermal transport modeling in nanomaterials, and advised senior design projects and independent study projects in similar areas.

  • [22]     M. P. Allen and M. P. Tildesley, Computer Simulation of Liquids. Oxford: Clarendon Press, 1987.
  • [23]     J. R. Lukes and H. Zhong, “Thermal Conductivity of Individual Single-Wall Carbon Nanotubes,” J. Heat Transfer, vol. 129, pp. 705–716, 2007.
  • [24]     M. B. Zanjani and J. R. Lukes, “Phonon Dispersion and Thermal Conductivity of Nanocrystal Superlattices using Three-Dimensional Atomistic Models,” J. Appl. Phys., vol. 115, p. 143515, 2014.
Molecular Diagnostic Sensors of Infectious Diseases at the Point of Care (Haim H. Bau, Mech. Eng) - bau@seas.upenn.edu

Website http://bau.seas.upenn.edu/

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 example, one of our recent projects focused on detecting the Zika virus in body fluids [25].  Our devices receive body fluids and carry out all the necessary operation to detect target nucleic acids associated with pathogens.  The devices have extensive multiplexing capabilities and interface with smartphones for signal acquisitions, processing, and reporting.  The project requires expertise from all engineering disciplines.  Participating students will be assigned mini projects compatible with their interests and abilities and our needs. The student will be exposed to challenging, cutting edge research problems, learn the basics of nanofluidics, microfabrication, sensing principles and electron microscopy.
Throughout the years, the laboratory has hosted many undergraduate researchers.  As an example, a recent bioengineering undergraduate Shih C. Liao has designed, constructed, and tested a diagnostic system that does not require electrical power [26].  The device is heated with an exothermic reaction of the type used in ready to eat meals and temperature is regulated with a phase change material.

  • [25]     J. Song*, M. G. Mauk, B. A. Hackett, S. Cherry, H. H. Bau, and C. Liu, “Instrument-Free Point-of-Care Molecular Detection of Zika Virus,” Anal. Chem., p. acs.analchem.6b01632, Jun. 2016.
  • [26]     S.-C. Liao*, J. Peng, M. G. Mauk, S. Awasthi, J. Song*, H. Friedman, H. H. Bau, and C. Liu, “Smart cup: A minimally-instrumented, smartphone-based point-of-care molecular diagnostic device,” Sensors Actuators B Chem., vol. 229, pp. 232–238, Jun. 2016.

Using Cell Phone Cameras, Optofluidics, and Signal Processing to Miniaturize Ultrasensitive Medical Diagnostics (David Issadore, BE and ESE) - issadore@seas.upenn.edu

Website: issadore.seas.upenn.edu
Droplet-based assays — in which ultra-sensitive molecular measurements are made by performing millions of parallel experiments in picoliter droplets — have generated enormous enthusiasm due to their single molecule resolution and robustness to reaction conditions [1]. These assays have enormous untapped potential for point of care diagnostics but are currently confined to laboratory settings due to the instrumentation necessary to serially generate, control, and measure tens of millions of droplets. To address this challenge, we are developing the microdroplet Megascale Detector (µMD), a new approach that can generate and detect the fluorescence of millions of droplets per second (1000x faster than conventional approaches) using only a conventional cell phone camera. The key innovation of our approach is borrowed from the telecommunications industry, wherein we modulate the excitation light with a pseudorandom sequence that enables individual droplets to be resolved that would otherwise overlap due to the limited frame rate of digital cameras. By miniaturizing and integrating droplet based diagnostics into a handheld format, the µMD platform can translate droplet based assays into a self-contained platform for practical use in clinical and industrial settings. This project is focused on understanding the operation of these microdroplet based biosensors and developing manufacturable fabrication methods to produce them. Work in this area will include custom electronics, programming in MATLAB, cloud computing, soft lithography, and microfluidic testing.
Ref 1: Muluneh, Melaku, et al. "Miniaturized, multiplexed readout of droplet-based microfluidic assays using time-domain modulation." Lab on a Chip 14.24 (2014): 4638-4646.

Touch sensors for brain-machine interface applications (Prof. Timothy Lucas, Neurosurgery; Prof. Jan Van der Spiegel, EE) -  jan@seas.upenn.edu

Website: http://sensor.seas.upenn.edu/index.php?title=Main_Page

The objective of this project is to identify and characterize implantable or wearable sensors of force and vibration at the fingertip. This project is motivated by the critical need to provide the sense of touch to paralyzed users of brain-machine interfaces (BMIs) to restore hand function. The work involves an initial comparison of the theoretical power, resolution, and readout requirements for the myriad transducers available to select those most unobtrusive to a user and operable by our recently developed wireless BMI electronics [28]. The second step involves characterizing the selected sensor(s) performance with bench and in situ testing. Specifically, the students will learn the theory behind various force transduction strategies and the power and communication requirements to deploy these sensors on the human hand. Prof. Lucas has mentored 8 high school and undergraduate students over the past 4 years, whose work has resulted in two published papers. Prof. Van der Spiegel has mentored over 45 undergraduate students many of which went on to graduate school. One of the recent papers with an REU student as co-author received the “Best Paper Award of the BIOCAS track” of the 2014 IEEE ISCAS conference [29].

  • [28]     X. Liu, M. Zhang, B. Subei*, A. G. Richardson, T. H. Lucas, and J. Van der Spiegel, “The PennBMBI: Design of a general purpose wireless brain-machine-brain interface system,” IEEE Trans. Biomed. Circuits Syst., no. 9, pp. 248–258, 2015.
  • [29]     X. Liu, B. Subei*, M. Zhang, A. G. Richardson, T. H. Lucas, and J. Van Der Spiegel, “The PennBMBI : A General Purpose Wireless Brain-Machine-Brain Interface System for Unrestrained Animals,” in 2014 IEEE International Symposium on Circuits and Systems (ISCAS2014), 2014, pp. 650–653.
Sensors for Tactile Feedback for Bilateral Activities of Daily Living Exercise Robot (Bi-Adler) - (Prof. Michelle Johnson, Med. Sch. & Bio Eng.) - johnmic@mail.med.upenn.edu

Website: http://www.med.upenn.edu/rehabilitation-robotics-lab/

We have developed Bi-ADLER, which is a therapy robot designed to help treat neurological disorders such as stroke and cerebral palsy. It is specifically designed for patients with upper extremity impairments, wherein one arm is impaired and the other one has higher degree of functionality.  The active robot arm is mainly for supporting the impaired arm and the passive arm is controlled by the patient's unimpaired arm.  Reach and grasp is critical for patients after a stroke to relearn both unilateral or bilateral coordination for activities. Non-invasive tactile sensors that are able to provide grasp force feedback without obstructing subject's ability to grasp and sense objects are required. Students will develop these sensors that can integrate seamlessly with the BiADLER robot to facilitate relearning after a stroke. Prof. Johnson is active in mentoring minority students and has several papers with undergraduate students as co-authors. [30] [31][32][33]

  • [30]     C. Lott* and M. J. Johnson, “Upper Limb Kinematics of Adults with Cerebral Palsy on Bilateral Functional Tasks,” in 38th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 2016.
  • [31]     E. Dimwamwa* and M. J. Johnson, “Kinematic analysis of unilateral and bilateral drinking task after brain and periphery injuries,” in 2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 2015, pp. 4558–4561.
  • [32]     D. . Adewole* and M. J. Johnson, “A computer model of the human arm: Predictive biomechanics forthe theradrive rehabilitation system,” in 2015 IEEE International Conference on Rehabilitation Robotics (ICORR), 2015, pp. 798–803.
  • [33]     R. Wilk* and M. J. Johnson, “Usability feedback of patients and therapists on a conceptual mobile service robot for inpatient and home-based stroke rehabilitation,” in 5th IEEE RAS/EMBS International Conference on Biomedical Robotics and Biomechatronics, 2014, pp. 438–443.
Non-invasive Detection of Interaction Forces During Patient-Therapist Rehabilitation - (Prof. Michelle Johnson, Med. Sch. & Bio Eng.) - johnmic@mail.med.upenn.edu

Website: http://www.med.upenn.edu/rehabilitation-robotics-lab/

By 2030, there will be about 10.8 million older adults living with a disability due to stroke. 75% of adults recovering from stroke have impairments in their upper-limb. Strong evidence supports the fact that rehabilitation robots can play a role in improving functional outcomes after stroke. They can supplement the shortage of rehabilitation therapists, provide more therapy time, and improve the cost-effectiveness and efficiency of therapy teams.
It is however unclear how these robots should behave in order to provide cooperative and collaborative therapy when interacting in a robot-patient dyad. While we are able to use inertial sensors to measure kinematic signatures, there is a need for the development of non-invasive tactile sensors that are able to allow researchers to measure the interaction forces being exchanged during these patient-therapist dyads. Our working hypothesis is that by studying kinematic and kinetic signatures arising from interactions of real therapist-patient dyads during neuro-rehabilitation, we can determine physical behaviors needed for a low-cost, therapy robot to function effectively in therapy settings.  The goal of this project is to investigate the development of such tactile sensors that are wearable, small, and invisible in that they do not interfere with the process of therapy.

Attitude sensor for novel flying devices (Prof. Mark Yim, Mech. Eng) -  yim@seas.upenn.edu

Website: http://modlabupenn.org/

Unlike common quadrotors popular today, dual rotor and single rotor devices are being explored [34].  This proposal is part of the “experimental flyer” project that consists of a small and lightweight device that is capable of controlled flight. The project focuses on the development of a state estimation system for the flyer that allows for precise and rapid attitude measurement.  An attitude sensor that measures a vehicle's attitude (global orientation) requires a three-axis magnetometer, a three-axis accelerometer, and three one-axis rate gyroscopes [35]. The students will filter and fuse sensor measurements together using a Square-Root Unscented Kalman Filter (SR-UKF) developed in MATLAB. The student will also study the feasibility of porting the filter to small microcontroller. This project will be suitable for a student in Mechanical, Electrical or Computer engineering.  He/she will learn MATLAB, circuits & microcontrollers, dynamics and control.

Professor Yim has been supervising over 20 undergraduate students and successfully integrated them in meaningful research projects. For example, the idea for this project was initiated by another REU student (C. Baldassano) who went on for his PhD at Stanford, and is currently a postdoc at Princeton University.

  • [34]     J. Paulos and M. Yim, “Flight performance of a swashplateless micro air vehicle,” in 2015 IEEE International Conference on Robotics and Automation (ICRA), 2015, pp. 5284–5289.
  • [35]     J. Paulos and M. Yim, “An underactuated propeller for attitude control in micro air vehicles,” in 2013 IEEE/RSJ International Conference on Intelligent Robots and Systems, 2013, pp. 1374–1379.
Robot with biologically inspired gaits (Mark Yim, Mechanical Eng.), yim@seas.upenn.edu

Website: http://modlabupenn.org/

The goal of this project is to explore tuning biologically inspired robot gaits for rapid and efficiently locomotion [1] . Doing so manually is a time-consuming and tedious task. For this project, CKBot (a modular robot system) will be configured into a centipede-inspired configuration with six legs.  The project requires understanding the parameters that effect gaits, which follow an alternating tripod pattern. The purpose of this project is to automate tuning of the robot's gait using motion capture sensors such that it optimizes a factor such as specific resistance, speed, or power. Automatic optimization technics such as Nelder-mead will be employed.

The student will set up multiple test scenarios including creating and building sensor frames and infrastructure, the Matlab optimization code, and a power regulators.  In addition, power monitoring sensing circuitry and telemetry need to be integrated with the MATLAB testing code. This project was started by another REU student (Sarah Koehler) who went on for a PhD at UCB.

  1. K. C. Galloway, J. E. Clark, M. Yim, and D. Koditschek, “Experimental investigations into the role of passive variable compliant legs for dynamic robotic locomotion,” in Robotics and automation (icra), 2011 IEEE Intl. Conf. on, 2011. doi:10.1109/ICRA.2011.5979941
Understanding and Exploiting Data from Hyperspectral Cameras for Agriculture -  cjtaylor@seas.upenn.edu

Website: http://www.cis.upenn.edu/~cjtaylor/RESEARCH/research.html

Over recent years the cost of purchasing hyperspectral cameras that are capable of measuring the spectrum of incoming light have dropped in price. Some companies are currently retailing versions of these systems that can operate in a push broom manner which allows them to be used in conjunction with GPS guided unmanned aerial vehicles to acquire ortho rectified images of terrestrial spaces. This project will explore how current data analysis techniques and classification systems could be used to analyze the incoming hyperspectral mosaics to better monitor the health of crops and to separate crop canopy from other vegetation in the field. Students will get familiar with spectrum specification and its relationship to crops and other vegetation, as well as various analysis techniques. Prof. CJ Taylor has been mentoring many undergraduate students over the past twenty years and has co-authored several publications with undergraduates.

An Aerially Deployable Environmental Sensor Probe (Prof. Vijay Kumar, Mech. Eng.) – kumar@seas.upenn.edu

Website: http://www.kumarrobotics.org/dr-vijay-kumar/

In precision agriculture and environmental monitoring, there is a need for physical sample acquisition for ex-situ analysis. In plant pathology applications for example, it is essential to collect samples of leaf, air, soil, etc. for studying disease onset and spread, enabling timely mitigation. The GRASP Lab is developing an environmental sensor probe that can be deployed and recovered autonomously by an aerial robot. The current prototype is a pest-trap in connection with USDA funded research for pest-density monitoring, whereas future versions will facilitate collection of fungal spores and other samples of interest. Prof. Kumar's group has a long tradition of mentoring undergraduate researchers. E.g. Delaney Kaufman,a rising sophomore at Penn, has been working on this project through summer 2016. Another undergraduate (Anurag Makineni) is a co-inventor on an agriculture US patent (pending). In addition, the work of undergraduate students often leads to publications (IEEE CASE 2015, IEEE ICRA 2015).
The REU student will familiarize themselves with the architecture of the aerial robot, study the probe prototype design, and help extend it for improved autonomy as well as energy sustainability (onboard solar cells). During the course of the project, the student will have the opportunity to work with plant pathologists and agronomists. The student will also participate in regular field experiments (on and off-site), and have the opportunity to work with other undergraduate and graduate students.

Estimating the energetics of aerial robot flight
kumar@seas.upenn.edu; Dr. Konstantinos Karydis; kkarydis@seas.upenn.edu

Websites: https://sites.google.com/site/kkaryd/  and http://www.kumarrobotics.org/dr-vijay-kumar/

Recent advances in aerial robotics and quadrotors specifically have created several opportunities in practical applications.  Search-and-rescue, precision agriculture, environmental monitoring, and airborne photography are just a few applications that have benefited from the improvement in our understanding on how to create motion and control quadrotors, as well integrate sensory feedback to allow for (semi-) autonomous operation in real-world settings.  However, the limited time these vehicles can stay airborne remains a key challenge.  One way to address this challenge is by rendering aerial robot operation more energy efficient.  To this end, the Multi-Robot Systems Lab (MRSL) at the University of Pennsylvania (UPenn) is investigating the energetics of aerial robot flight.  Currently, the energy-efficiency of a prominent motion planning methodology---generating minimum snap trajectories---is evaluated against other methods such as minimum kinetic energy trajectories.  Two undergraduate UPenn students,Nadia Kreciglowa and Shreetej Reddy were working on this project through Summer 2016.

The REU student will have the opportunity to learn the principles of multi-rotor aerial robot motion planning and control through performing mathematical analysis of the energetics of flight, and familiarize him or herself with machine learning to consolidate the developed theory.  The student will also learn how to control robots in simulation through the Robot Operating System (ROS), and will acquire hands-on experience by conducting data collection using motion capture systems.  Working with the researchers at MRSL will be a positive and fruitful experience for the REU student.  The lab regularly mentors undergraduate and high-school students, often leading to publications.  The postdoctoral researcher involved in this project is currently mentoring one master's and three undergraduate students at UPenn.  He has also gained extensive experience in engaging undergraduate students in research during his doctoral studies.
Dr. Konstantinos Karydis; kkarydis@seas.upenn.eduhttps://sites.google.com/site/kkaryd/

On-Chip Introspective Sensing and Adaptation (Prof. Andre' DeHon, EE) - andre@seas.upenn.edu

Website: http://ic.ese.upenn.edu/

Modern Deep Sub-Micron ICs are like snowflakes, every chip is different, and their characteristics change over time.  Consequently, it is valuable to enable chips to introspect on their behavior (e.g. delays) and self-adapt to their unique characteristics.  SUNFEST student T. Linscott measured on-chip delays and identified the effects of self-heating from clocking resources and the challenges that presents to on-chip delay measurement [36]. Taking this a step further, we've developed lightweight sensors for link delay measurement during operation coupled with fast repair strategies using pre-computed alternatives [37]. Based on these early results, we are interested in exploring a broad set of options for on-chip sensing and adaptation to improve reliability, extend lifetimes, reduce energy consumption, and increase performance.    Students working on these projects learn about the challenges of modern, nanometer-scale fabrication technologies, including variation and aging. Students will learn about modern reconfigurable architectures, lightweight measurement techniques, adaptive algorithms, and quantitative evaluation of costs and benefits.  The work exercises experimental methodology from formulation, through experiment design, experiment management, data collection, visualization, modeling, and explanation.
The Implementation of Computation Group has a long history of engaging undergraduate students in research, several of whom are co-authors and went on to graduate school [38], [39], [40], [41].

  • [36]     E. Kadric, P. Gurniak*, and A. DeHon, “Accurate Parallel Floating-Point Accumulation,” in Proceedings of the IEEE Symposium on Computer Arithmetic, 2013, pp. 153–162.
  • [37]     A. Kwon*, U. Dhawan, J. M. Smith, T. F. Knight, and A. DeHon, “Low-fat pointers,” in Proceedings of the 2013 ACM SIGSAC conference on Computer & communications security - CCS '13, 2013, pp. 721–732.
  • [38]     E. Kadric, K. Mahajan*, and A. DeHon, “Energy Reduction through Differential Reliability and Lightweight Checking,” in Proceedings of the IEEE Symposium on Field-Programmable Custom Computing Machines, 2014.
  • [39]     E. Kadric, K. Mahajan*, and DeHon. Andre', “Kung Fu Data Energy---Minimizing Communication Energy in FPGA Computations,” in Proceedings of the IEEE Symposium on Field-Programmable Custom Computing Machines, 2014.
  • [40]     A. Kwon*, K. Zhang, P. L. Lim, Y. Pan, J. M. Smith, and A. DeHon, “RotoRouter: Router Support for Endpoint-Authorized Decentralized Traffic Filtering to Prevent {DoS} Attacks,” in Proceedings of the IEEE International Conference on Field-Programmable Technology, 2014.
  • [41]       E. Kadric, D. Lakata*, and A. DeHon, “Impact of Memory Architecture on {FPGA} Energy Consumption,” in Proceedings of the International Symposium on Field-Programmable Gate Arrays, 2015, pp. 142–155.
Graph theoretic analysis of elections in the United States (Prof. Alejandro Ribeiro, Electrical and Systems Engineering) - aribeiro@seas.upenn.edu

Website: https://alliance.seas.upenn.edu/~aribeiro/wiki/

In the spirit of extracting data from unconventional sensing, this project uses sophisticated network theory and mathematics to study election outcomes. The first real uses of big data analysis and network theory to predict outcomes and direct campaign resources was in 2008 with the election of sitting President Barack Obama. This election saw Obama's campaign team leverage mathematics and big data to efficiently allocate campaign resources and also witnessed journalist Nate Silver use advanced statistical modeling to accurately predict election returns. Since 2008, political operatives of all camps have attempted to gain a better understanding of mathematics in order to replicate the success of the Democrats in 2008. Meanwhile, polling organizations are always attempting to refine their methodology and technique to more accurately predict elections.


This specific research project will aggregate historical presidential election returns at the county level with the intent of producing a correlation matrix that can be used to predict future election outcomes based on polling data. Using new research in graph signal processing the student will attempt to answer questions such as “Does the structure of the network facilitate prediction of election outcomes based on limited polling data?” and “How far in history should the network of counties reflect?” Conclusions to this research could be very valuable to polling organizations, news organizations, campaign teams, and other researchers.


Specific tasks given to the undergraduate include modeling and analyzing the presidential election network. The student will use a programming language (Matlab, Python, etc) to model the U.S. presidential election given the data provided. Models will be based on concepts such as probability theory, stochastic systems, and network theory. The undergraduate will also be comparing the results of these models to form conclusions on the effectiveness of polling data in predicting elections. 

Variability of Brain Activity during Learning (Alejandro Ribeiro, Electrical and Systems Engineering) - aribeiro@seas.upenn.edu

Website: https://alliance.seas.upenn.edu/~aribeiro/wiki/

The human brain is a complex system, extensively researched across multiple paradigms. Analysis of brain activity on a temporal and spatial scale can uncover information about patients in regards to learning abilities, neuronal defects, etc. It has recently become of interest to model neuronal activities using networks and signals supported on networks. Brain networks represent the levels of functionality between regions, and brain signals determine the level of regional brain activity. Signal processing techniques can therefore be used to map brain signals to a representative brain network, and examine brain activation patterns.
This project will entail analysis of functional MRI data using various graph signal processing techniques. Specifically, the student will decompose brain signals into their graph frequency representations, and analyze how activity patterns change both over time, and across different brain regions. The student should consider questions such as, “How do low and high graph frequency components differ?” and “How do these decomposed signals relate to the brain network?” The methods determined by the student will be applied to structural brain networks as well as fMRI data for both subjects conducting learning-based tasks, and subjects during their resting state.

The use of a programming language is required for the project, so that the student can process relatively large files and multiple types of data. Applications will be based on the student's ability to understand how to represent signals in its different forms, and consequently visualize how these forms are related on a brain network. Basic concepts of statistics and graphing techniques will be utilized and implemented. Further applications of this project could be to distinguish between subjects with Alzheimer's disease, or other neuro-degenetive diseases including Alzheimer's and Parkinson's.