Current Student Research Projects for 2017

Evaluation of Flattened vs. Round Preston Tube

  • Student(s): Samuel Hall Watson
  • Faculty: Russell V. Westphal
  • Report: Click to download

A study has been conducted to evaluate the use of flattened tubes for use in measurement of skin friction using the Preston tube method. Experiments were conducted in the Cal Poly 2x2 Foot wind tunnel wherein measurements from conventional, round, tubes were compared with flattened tubes. Ten different flattened tubes and ten different round tubes were employed. The results showed that the flattened tubes tend to measure skin friction values about 10% lower than round tubes when both use the same calibration relations. Hence, an alternative calibration equation is recommended for flattened tubes.

Aerodynamic Properties of a Representative Automotive Panel Gap

  • Student(s): Alejandro Meraz and Geary Yu
  • Faculty: Graham Doig
  • Report: Click to download

Within the automotive industry, there are uncertainties with regards to variations in panel gap width and depth, including their effect on wind tunnel and CFD models. An understanding of automotive panel gaps and their effects on aerodynamic properties will help Tesla engineers understand the impact of panel tolerances on the aerodynamics of a vehicle. Throughout the summer, a Tesla model was placed in a low-speed wind tunnel to understand the aerodynamic characteristics of different panel configurations such as an offset of the panel (-1, 0, and 1mm) and cavity volume (small and large). The Tesla model is a physical representation of a generic panel gap on an automobile. All wind tunnel tests were performed at a Reynolds Number of 3.2 million to mimic highway conditions. Surface pressure measurements indicated little differences in the near-wall flow apart from the center of the model, however wake profiles showed a larger discrepancy indicating more momentum loss. With the presence of a panel gap, boundary layer development was affected downstream. Tuft, paint, and smoke flow visualization tests were conducted to guide measurement locations and visualize flow separation on the model. Wind tunnel results were compared to CFD simulations as a form of validation. Based on experimental results, the inclusion of automotive panel gaps may not be necessary for wind tunnel and CFD models but will affect the accuracy of aerodynamic calculations.

Inkjet printed electrochemical, organic field-effect transistors

  • Student(s): Hajime Yamanaka, Juan Alejandro Ortiz Salazar
  • Faculty: Dr. Linda Vanasupa
  • Report: Click to download

A conventional inkjet printer was used to reproduce an electrochemical, organic transistor (ECT) on photopaper. Poly(3, 4-ethylenedioxythiphene):poly(styrenesulfonate), or PEDOT:PSS, and electrolyte went through an oxidation reaction to perform as a logical inverter. This mechanism is different from the conventional semiconductor transistor. The transistor was designed as a normally-on field effect transistor. The output from the circuit was square waves which had an amplitude of 0.4 V. This result was lower in magnitude than what is expected from the literature. The reason could be accounted for a slow switching speed at the channel of the transistor.

Reducing Electricity Consumption in Hybrid Residential System Using MISO Converter

This project investigates load distribution at a residential scale, comparing the efficiency of conventional models to those that have a separate circuit for DC loads at different ratios. The distribution models include: (1) AC generation and AC load distribution, (2) AC generation and hybrid AC/DC load distribution, (3) AC and inverted DC generation with AC load distribution, and (4) AC and DC generations with hybrid AC/DC loads incorporating Multiple Input Single Output DC-DC converter (MISO). Results demonstrate that across several ratios of DC loads, the hybridization of load distribution increases the overall system efficiency; thus, reducing electricity consumption. Economic analysis of the results further suggest that adopting renewable generation provides significant financial benefit in terms of return on investment, and that hybridizing the load distribution increases the benefits, even with a slightly higher initial investment.

Development of New Laboratory Experiments for the Combustion Engine Design Course

  • Student(s): Dorian Capps and Michael Bolton
  • Faculty: Dr. Patrick Lemieux

The 2017 Summer Undergraduate Research Program for the Engines and Propulsion Laboratory resulted in the development of two new laboratory experiments which students taking ME 444 can perform. The first lab experiment is a vibrational analysis lab where students will examine the natural frequencies of two single-cylinder diesel engines made by Hatz. The two Hatz engines are similar in every way except that one features balancing shafts while the other does not. Using a Magtrol dynamometer, triaxial accelerometer, and National Instruments data acquisition setup, students can obtain torque, horsepower, and vibration data all at once. They then swap one engine out for the other and again collect data to compare the two engines and validate theoretical calculations. For students to smoothly swap engines and collect all necessary data within the allotted time for a lab, each engine required its own mounting system which could accurately locate the engine’s output shaft to the dynamometer’s input shaft quickly and consistently. This was accomplished by using locating pins and fixture keys to precisely fasten the dynamometer to a slotted table. The engines and their mounts, which also feature fixture keys, are then lowered onto the table and slid forward along the slots to mate with the dyno coupler. Distortion of the engine vibration data is avoided by isolating the engine from its mount with rubber. A dyno coupler had to be specified with a large enough misalignment tolerance to accommodate the engine’s displacement that results from the measured vibrations. The second lab experiment which saw progress was for the JFS-100 power turbine. A modification to the fuel flow governor was achieved and improvements to the controller software were made so that several test fires were conducted and the collected data is being used to guide the continued advancement of this lab.

Identifying Perils of Mining Github Data

  • Student(s): Max Moede
  • Faculty: Davide Falessi

Problem context: Mining data from GitHub repositories resulted to be useful for validating several software engineering theories such as defect prediction and code smells. However, no repository is perfect and it is important to understand which repositories are valuable to mine. My main goal was to analyze and characterize the validity level of GitHub repositories. Challenges: One of my major challenges was determining which metrics could be calculated with available Python libraries. I had very little experience with Python before this project, so I had to learn the language and learn how to use libraries like PyGitHub and GitPython. This was also the first time I had to work remotely, which limited my ability to communicate with my team. Results: We successfully gathered metrics for about 200 Apache Software Foundation projects, the majority of which passed as valid projects. A few of the older apache projects were deemed invalid mainly due to inactivity and unreliable commit data. Learned Topics: I have learned how to write effective scripts in Python and developed a thorough understanding of the GitHub API. I also learned how to read API documentation in a practical way. I was able to gain experience with databases and using MySQL as well. I have also learned why it is important to select specific repositories to perform reliable software engineering research. Next Steps: The project lead to an independent study (CSC 400). I am currently attempting to develop a website which would allow users to analyze software projects of their choice. I am also writing a paper about the results of the projects which we plan to submit in a major internal peer-review conference such as the Mining Software Repository conference:

Perspectives on Managing Emergent Risk due to Rising Complexity in Aerospace Systems

  • Student(s): Andrew M. Palmer
  • Faculty: Dianne J. DeTurris

Managing complexity in aerospace is essential for effective modern systems development. Preventing unexpected or emergent risk while maintaining cost and schedule discipline is the main objective of addressing unwanted self-organization due to complexity. The emergent risk can be quantified by viewing systems from multiple stakeholder and design perspectives to accurately assess unforeseen behaviors. Novel models of systems design and analysis from other industries and disciplines are introduced and placed in an aerospace context. This survey highlights commonalities between complexity management methods, models of systems design, and analysis that can be used to predict and prepare for emergent behavior in the system. Methods for dealing with complexity include systems thinking, complex adaptive systems, the Cynefin framework, Agile principles, managing complexity risk, and complexity measurement. The methods emphasize systems-thinking using stakeholder perspectives as a mechanism to minimize emergent risk. We also provide an integration of frameworks, within the context of systems engineering, as applied to design and lifecycle concerns. Some or all of these methods can be used simultaneously to provide new perspective on a given aerospace system. New structures for system engineering are necessary to aerospace as a continuing “bureaucracy of innovation” to prevent catastrophic failure due to overwhelming complexity. Using new points-of-view to design gives practicing engineers more information with which to view and manipulate system properties. The aerospace engineering process requires new perspectives on rising complexity in order to meet the challenges of modern engineered systems. The solution is a collaboration between disciplines in order to solve problems that span multiple academic domains.

Enhancing Software Impact Analysis via Semantic Requirements Analysis—SURP 2017

  • Student(s): Tina Rickard, Tyeler Bridges, Max Moede, Michael Lozada, Yiupang Chen
  • Faculty: Davide Falessi

Context & Aim: When a new requirement is introduced into a piece of software, the developers must identify the impact set—the set of source code classes that would be affected by implementing this requirement. This practice, called change impact analysis, becomes more difficult as the code grows or new developers join. My aim in this project is to develop an open-source software that aids developers in the three areas of change impact analysis: 1) identifying an accurate impact set, 2) deciding where to refactor, and 3) planning a project. Challenges, Results, & Lessons: One of the biggest challenges was learning different software. Specifically, I have revisited Python and had to learn from scratch the GitPython, SciPy, and Genesim libraries. In the interest of time I initially searched the internet for solutions to specific problems. However, I quickly learned that while this provided a temporary fix, it was better to actually learn the language or API so I would not have to constantly search for different solutions. I have often collaborated with my classmate Tyeler Bridges, who was also part of the project; a challenge we had was scheduling meetings given our different schedules. We were not able to meet very often or for long, and our communication was sometimes staggered. As we were working largely separately on our task, we had issues in putting our finished parts together, often needing clarification on the inputs and outputs of our programs. However, as of this writing we have made considerable progress together and have developed a beta release of a program that calculates certain code metrics for the commits over a requirement’s implementation. Next Steps: I will continue to work on this project as part of an independent study, CSC 400. The immediate goal for me is to finish the metrics calculator and following this, polish and improve the project. I am also writing a paper about the results of the projects which we plan to submit in a major internal peer-review conference such as the Mining Software Repository conference (

Autonomous Vehicle Development

  • Student(s): Tristan Perry and Paul Rothhammer-Ruiz
  • Faculty: Charles Birdsong

This research is towards constructing a fully-operational small-scale intelligent vehicle. The small-scale intelligent vehicle will be used as the platform for Dr. Birdsong’s advanced controls course. The vehicle is built from RC car components and is equipped with a Raspberry Pi and a Teensy microcontroller. Part of the work this summer focused on establishing fault proof communication between the Raspberry Pi and the microcontroller. Another part of the work this summer was to build a vehicle model of the steering to allow for prediction of vehicle performance in various environments. With this vehicle model, MIMO control algorithms were then designed to simultaneously control the speed and steering of the vehicle. In order to allow for feedback to the controller, an IMU sensor was integrated to measure speed and heading angle of the vehicle. In addition to the IMU sensor, a camera sensor was used to detect lane lines to allow for lane keeping. The work done this summer allows for the Raspberry Pi to run control algorithms that can simultaneously control steering and speed of the vehicle.

Wind Tunnel Model Creation Using a Biomimetic Approach

  • Student(s): Graham Rolph, Stavros Diamantopoulos
  • Faculty: Dr. Graham Doig
  • Report: Click to download

This summer, research was undertaken to further expand upon previous thesis work done by formar graduate student David Martin. Utilizing guidance from David’s Report, a white pelican taxidermy, courtesy of the Cal Poly Biology Department was 3D scanned at NASA Ames Research Laboratory and converted into 3D models of one wing and half of the body using MeshLab. The models were then adjusted so that they accurately mimic the wing of a white pelican in flight using Blender. Once an accurate model was created, it was scaled to 80% its original size, so that the wing could be positioned in the tunnel. The model was then partitioned into multiple sections so that they can be 3D printed using the Lulzbot Taz 5 and Taz 6 printers in lab. The primary feathers, due to their shape, were printed using Taulman3D T-Glase PET filament and the rest of the wing was printed utilizing High Impact Polystyrene. Once the model was printed and assembled, the parts were sanded and joined using Bondo and epoxy. It was then mounted in Cal Poly’s Low Speed Wind Tunnel and tested across a range of conditions to ensure its stability during testing. Future research using smoke and laser flow visualization will qualitatively locate the wingtip vortices. Testing utilizing a total pressure probe and rake behind the wingtip will be able to quantify the strength of the vortices.

A Study of Financial Management and Accounting Techniques in Startups and Entrepreneurial Ventures

A study of financial management and accounting techniques in startups and entrepreneurial ventures was conducted by Jason A. Chang, a masters student from California Polytechnic State University, San Luis Obispo. Michelle Wong worked closely with Professor Liz Schlemer to analyze 13 interviews with Entrepreneurs. The financial management and accounting practices used in the 13 startups were analyzed in order to find trends in the financial methodologies. The Balance Sheet, Income Statement, Statement of Stockholder’s Equity and Cash Flow were the four financial statements analyzed. The success metrics defined by the startup were also observed. A framework outline was outlining the differences in financial tools based on the maturity of the enterprise was developed. A draft paper was written in preparation for publication in 2018. Early startups are unable to use the four financial statements for its intended purpose because the businesses were not mature enough. Instead, the early startups use an abridged version of the four financial statements. On the other hand, late stage startups use the four financial statements in a traditional manner. A common success metric seen across all startups is business development. Business development encompasses user growth, obtaining funding, and improving bottom line numbers.


  • Student(s): Jalen Mano
  • Faculty: Dr. Eltahry Elghandour

Present day composite sandwich panels provide incredible strength. Their largest problem, however, is early bonding failure between the core and the skin. This is usually due to the low bonding surface area of present cores like honeycomb. Corrugated structures could provide a remedy for this with their much larger bonding surface area. Corrugated structures have extreme mechanical properties deeming them particularly useful in aerospace and automotive applications. Bio-composites have properties that could strengthen the corrugated sandwich panel against the delamination of the skin and increase the strength of the structure while making it cheaper and more environmentally friendly. This paper presents the optimum design, manufacturing, and testing of corrugated sandwich panel structures under compression loading. To do this, optimum corrugation geometry identified using theoretical analysis of the moment and bonding area of the shape. A bio-composite was integrated in both the core and the skin individually in corrugated sandwich panels. The cases tested were all carbon fiber, hemp skin with carbon fiber core, carbon fiber skin with hemp core, and all hemp. These corrugated structures were analyzed by conducting compression loading tests on varying lengths of single-ligament panels utilizing trapezoidal corrugation as the core and a flat plate as the skin. For these test samples, the optimal shape and length of the corrugation was investigated. These both impact the bonding surface area and bonding strength where the material was most subject to failure. Lengths were systematically tested to determine the effect the length had on the bonding shear strength and buckling behavior. Test results showed the bio-composite acting as a skin and a core had a significant effect on the mechanical behavior of the panel under a compression load while the length had a significant effect on the bonding failure modes.

Bio-remediation of MTBE and TBA in Groundwater

  • Student(s): Nilma Edward, Nancy Lam
  • Faculty: Dr. Nirupam Pal

Methyl-tert butyl ether (MTBE) is an oxygenate added to gasoline to improve the quality of air emissions. However, in water MTBE has been known to cause health issues such as headaches in high doses and is a suspected carcinogen. One of its byproducts, Tert-Butyl Alcohol (TBA), is also a known toxin which can cause reproductive and developmental defects. Both are highly soluble in water and difficult to degrade. Gasoline stored in leaking underground storage tanks has led to the contamination of groundwater wells throughout California. Biodegradation has been tested as a potential treatment option. To test its viability in both anaerobic and aerobic conditions, vials containing combinations of MTBE and TBA with groundwater bacteria were compared to positive and negative controls. The positive control was a mixture of TBA, MTBE, biowish bacteria, a mineral salt solution, and a growth media. The negative control was a mixture of TBA, MTBE, a mineral salt solution, and growth media. Gas Chromatography/ Mass Spectrometry and Solid Phase Micro-Extraction were used to analyze the concentrations of the samples every 48 hours. Spectrophotometry was used to determine the bacterial growth of each sample every 48 hours. The project results were inconclusive and further testing is required.

Cell Transplantation to Enhance Natural Bypass Enlargement Following Arterial Occlusion

  • Student(s): Isobelle Espiritu
  • Faculty: Trevor Cardinal

In 2015, 155 million people worldwide were affected by peripheral arterial occlusion disease (PAOD). This disease generally involves the occlusion of limb arteries through atherosclerosis and results in restricted blood flow and tissue oxygen delivery. Some patients have pre-existing collaterals that serve as natural bypasses around occluded arteries. These natural bypass collaterals can enlarge to preserve tissue health. Unfortunately, enlargement of collaterals is insufficient or impaired in many patients. Work in our lab uses mice to show that following arterial occlusion through ligation, capillaries are capable of arterializing to restore blood flow. To enhance this process, we transplant muscle stem cells, called myoblasts, into the ligation site of the target muscle. Previous work demonstrated that myoblasts transplanted on a gelatin carrier increased the amount of enlarged collaterals 7 days after ligation. A gelatin carrier was used rather than an injection because injections can be imprecise or cause trauma. However, these gelatin carriers had several limitations. The preparation process to create the gelatin and inoculate the cells was time-consuming and challenging. Additionally, once the cells were inoculated in the gelatin, they had to be incubated for several hours prior to transplantation. Long incubation times resulted in less effective cell therapy. Finally, the gelatin itself is bio-derived, therefore, presented a higher variability and required modifications. My research focuses on optimizing the current cell transplantation protocol by testing a synthetic, commercially available thermo-reversible hydrogel to allow the cells to be transplanted more efficiently. This would preserve cell effectiveness. To achieve this objective, I learned cell culture, surgical procedures for inducing arterial occlusion and transplanting cells, and began testing the hydrogel. Future studies will involve building upon previous students’ work and testing the new vehicle to transplant cells then observing the cell therapy effects at additional time points following the ligation surgery.


  • Student(s): Max Selna
  • Faculty: John Bellardo

QuickSat, once it is complete, will be an on-shelf 1U technical Demo that will serve as a substitute for flying mass models. It will enable PolySat’s experimental systems to gain flight heritage on a non-critical mission. This summer, I worked on developing a design for a pin-less Nitinol hinge that will deploy side panels with solar cells. These hinges use thin, broad sheets of nitinol to maintain the same restoring moment as a wire hinge design. This makes for a higher number of endurance cycles and a more dependable clamping interface with the structure. Regarding the material properties of the NiTi, I researched different grades and concluded that an Austenite transition temperature of -19C was necessary to ensure that the hinges would deploy with sufficient torque in a space environment. The Austenite transition temperature is where the lattice structure of a material changes. For most metals, such as steel, this temperature is far above ambient temperature, making use of the material properties in the austenite phase impractical. With NiTi, we can have the effects of a ductile material at low temperatures, and the effects of an elastic material at high temperatures, due to the phase change. The main concern now is that the hinges may be too cold, causing the NiTi being in the wrong phase. This could cause the hinge to not deploy. The beginnings of the ETU structure were machined in Mustang ’60 this summer, and the rest of the structure will be finished by the end of the quarter.

Automated Image Data Acquisition for Concentration Gradient Generating Microfluidic Chip

  • Student(s): Kevin Wohlfarth
  • Faculty: Benjamin Hawkins

The scope of my summer undergraduate research project was to assist in automating and developing processes utilizing various hardware for future use as a cell culture platform and for Dr. Hawkins to continue conducting his independent research. Tasks for this project included developing a fully automated data collection system with various hardware, learning soft lithography synthesis techniques and validating theoretical results via image analysis. The Labsmith uProcess programming language was used to control an inverted fluorescent microscope (LabSmith SVM340) to collect 16 images within a concentration gradient generating microfluidic chip for post image processing across the culture wells. Each microfluidic chip was manufactured onto a microscope slide using soft-lithography techniques within the Cal Poly Micro-fabrication Laboratory. Two syringe pumps were used to distribute blue and yellow dye at equal flow rates of approximately 0.5 ml/hr into the chip to generate a visible gradient of blue, green and yellow within each of 5 mm diameter wells. Image analysis included manipulating raw image data into total intensity values for each respective image to create plots that demonstrated the effectiveness of the concentration gradient. Future directions for the project include working with fluorescent stained cell culture and comparing image data with measured electrical impedance data across each culture well to observe cell growth when exposed to varying concentrations of various solutions.

Small Scale Intelligent Vehicle

  • Student(s): Paul Rothhammer-Ruiz, Tristan Perry
  • Faculty: Dr. Charles Birdsong

This research is towards constructing a fully-operational small-scale intelligent vehicle. The small-scale intelligent vehicle will be used as the platform for Dr. Birdsong’s advanced controls course. The vehicle is built from RC car components and is equipped with a Raspberry Pi and a Teensy microcontroller. Part of the work this summer focused on establishing fault proof communication between the Raspberry Pi and the microcontroller. Another part of the work this summer was to build a vehicle model of the steering to allow for prediction of vehicle performance in various environments. With this vehicle model, MIMO control algorithms were then designed to simultaneously control the speed and steering of the vehicle. In order to allow for feedback to the controller, an IMU sensor was integrated to measure speed and heading angle of the vehicle. In addition to the IMU sensor, a camera sensor was used to detect lane lines to allow for lane keeping. The work done this summer allows for the Raspberry Pi to run control algorithms that can simultaneously control steering and speed of the vehicle.

Aerodynamic Deorbit Experiment (ADE) CubeSat

  • Student(s): Liam Bruno, Arielle Cohen
  • Faculty: Dr. John Bellardo

ADE is a 1U CubeSat with a deployable drag sail payload. It will be deployed into a geostationary transfer orbit (GTO). The primary mission objective for ADE is to provide flight qualification for and demonstrate the viability of its payload. The deployable drag sail is designed to take advantage of the aerodynamic drag forces experienced by the spacecraft and decrease the time it will take to fully de-orbit. A successful demonstration of ADE’s payload would be a step closer to the mitigation of a growing problem in spaceflight—orbital debris. In addition to demonstrating an aerodynamic deorbit technology, ADE will attempt to characterize the radiation environment in GTO, and provide invaluable data for future small satellite missions in radiation-heavy trajectories. During the summer quarter, we primarily focused on understanding the limitations introduced by the radiation environment of GTO, and making appropriate design choices to minimize negative effects. After conducting analysis on the radiation levels that ADE will be exposed to, the decision was made to introduce additional shielding as well as a radiation sensor into the design of the flight electronics. As a result of this shielding, the system’s mass became excessive, and a re-design of the structure was performed to compensate. In general, advancements in the mechanical and electrical design of ADE were made throughout the summer in preparation for critical design review as well as further development and testing.

Improving Mechanical Performance and Manufacturing Techniques of Natural Composite Structures

  • Student(s): John Niemoller
  • Faculty: Eltahry Elghandour

Composite materials are one of the most widely used forms of structural reinforcement, and their application is seen across a variety of different industries. Traditionally, composites have consisted of inorganic fibers in petroleum based resin matrices, which are both hazardous to manufacture and difficult to recycle. However, bio-composites offer an eco-friendly solution through means of natural waste material for use as a structural reinforcement. Within the past several years, bio-composites have infiltrated many industries, and this report focuses on their impact for using them as a structural member and building material. The application of bio-composites for use as a structural member have been recently sought out due to their mechanical performance, availability, and environmental impacts. Of practical interest to this project, bio-composite sandwich structures were manufactured, compression tested, and analyzed for their mechanical properties, for any structural application. The sandwich structure was constructed from multi-function materials to utilize the compression load. The structures materials investigated consisted of hemp, cornhusk, epoxy, and silica aerogels. The manufacturing of these sandwich structures was done to be a single cell trapezoidal core with two flat composite sheets bonded on either side to provide added stiffness to the structure. Manufacturing these structures required many iterations in order to derive an effective technique. Three similar types of sandwich structures were built and tested. The first specimen consisted of using dried cornhusk for the core cells and flat sheets, with the fiber orientation in the load direction. The second specimen was constructed with a core containing hemp and cornhusk. The third specimen was constructed with a core containing hemp, cornhusk, and varying quantities of silica aerogels. The results from this project produced a technique for fabricating cornhusk based structures, and proved integration of silica aerogels in a multifunctioning material have a significant effect on mechanical performance.

Removal of Hexavalent Chromium in water using Magnetite Nanoparticles

The adsorption of chromium hexavalent in water by magnetic magnetite nanoparticles, synthesized by chemical coprecipitation, was tested using batch absorption experiments. The influence of time and dosage on the removal efficiency of chromium was investigated. High removal efficiencies were obtained for hexavalent chromium. The adsorption data obtained from the experiments were fitted to Freundlich and Langmuir adsorption isotherm models. The Langmuir model best described chromium adsorption on magnetite. This research shows that magnetite nanoparticles have a high potential for the removal of hexavalent chromium in water treatment as a result of high removal rates and potential for recyclability and reuse of the spent adsorption media.

Developing Wireless Vehicle-to-Vehicle Communication Systems in Collaborative Autonomous Vehicles

Collaborative Autonomous Vehicles are fast becoming an integral part of our 21st century mode of transportation. Though presently in an R&D phase, they have proven to be environmentally friendly and a safer way to travel by eliminating human error. The only major risk to autonomous vehicles is the risk of external attack where in which an attacker can remotely cause collisions to the cars in the platoon In this project, on the security vulnerabilities in autonomous vehicles. Work on the project was done in three phases, namely, autonomous, collaborative and testing. The autonomous phase was completed spring quarter of 2017. In this phase, we designed a vehicle to track an image in front of it, and based on the object position the car either moved forward or stopped. The collaborative phase of the project was completed summer quarter of 2017. In this phase I designed and implemented and wireless network through which the cars could communicate with one another. The testing phase should be completed by the fall quarter of 2017. In this phase, we are going to run a series of security tests on the platoon to identify vulnerabilities in the autonomous vehicles. During my SURP, I created a wireless network through which the cars could relay information to one another through a UDP socket implemented in python. The functionality I implemented included a router attached to one of the cars, which forwards packets on a wireless computer network. Some of the data relayed includes the distance and position value of the lead car and other cars in the platoon and in the future logs of internal functionality for each vehicle. Being passionate about computer networks, I felt that this project was a great fit for me and thus I intend to extend this work for my senior project.

Comparison of Keratinocyte Cell Lines

  • Student(s): Stephanie Switalski
  • Faculty: Lily Laiho

Cell culturing is the fundamental base off which cell and tissue-based research projects are built. The purpose of this summer research project was to determine if cells obtained by different vendors would behave differently. In this study, keratinocyte (KRTs) cell lines from American Type Cell Collection (ATCC) and Lifeline Cell Technology (LLCT) were assessed based on cell morphology and growth, and response to ultraviolet (UV) light. The cell growth patterns of LLCT were extremely variable, and therefore unpredictable. While cell growth rate of ATCC was not linear, it was predictable and repeatable. Keratinocyte cell lines are primary cells, which are cells that are isolated directly from tissue and contain both the cell of interest and other cells harvested from the tissue. As cells are passaged, other cells quickly die leaving only the cell of interest. After two passages, the ATCC line was purified to only KRTs. After four passages, the LLCT line still contained non-keratinocyte cells, making them non-ideal for experiments. In addition, the LLCT KRTs displayed poor morphology on passage 4, not displaying the characteristic “cobblestone” appearance of healthy KRTs. A sample of atypical morphology and non-keratinocytes in the LLCT cell line is available in the abstract image (blue: DAPI - nucleus, green: p21- UV damage). ATCC cells expressed typical KRT morphology until passage 6. In previous experiments, ATCC cells were exposed to UV and then cell damage was quantified using p21, a general damage marker, using fluorescence microscopy. The same experiment was performed using LLCT cells. Results were similar; however, there was higher variability within the cell response, which would affect results in any experiments conducted. In conclusion, cells obtained from different manufacturers can perform differently, leading to more variability in experimental results. The ATCC keratinocyte cell line provides more homogenous cell cultures to conduct the lab’s research.

Bicyclist behavior at Two-way Stopped Control Intersection: Implications for 'Idaho Stop Control'

  • Student(s): Avery Lai
  • Faculty: Anurag Pande

As an active mode of transportation, bicycling can provide significant benefits for the physical well-being of the riders in addition to the environmental benefits. In spite of the benefits, there are significant challenges to widespread use of bicycles in the US. Overcoming these challenges to increased bicycle usage requires that the needs of bicyclists on the road are adequately identified and addressed. In this regard, bicyclists in Idaho have been allowed by law to slow down at stop signs, check for cars and pedestrians, and essentially treat a stop sign as though it were a yield sign. A similar change in law has been proposed in California. Although the law may seem controversial, the so-called Idaho Stop has decreased bicycle injuries in Idaho by 14.5% the year after the law took place. As part of this research experience, we observed the bicyclist behavior at 2-way stop controlled intersection and found that the % of bicyclist actually making a complete stop is negligible. This is consistent with our anecdotal observations as well as the citation data obtained from the Cal Poly Police department. There have been 265 citations/warnings issued to bicyclist since January 1, 2016, on the campus alone. Campus police have put special signs for bicyclists to address this issue (See Image). Our conclusion from this work is that a change in law allowing bicyclists to legally make an ‘Idaho Stop’ would really be codifying the existing behavior. It may reset motorist expectations about what the bicyclist is going to do and may, therefore, have some safety benefits.

A social information managing game engine

  • Student(s): Liam Gow, Nathan Philliber, Zoe Cagle
  • Faculty: Foaad Khosmood

In this project, we designed an engine to facilitate a massively multiplayer online role playing game (MMORPG) using information trade/acquisition as the primary mechanic. While most MMORPGs have physical goals such as annihilating foes or obtaining important objects, the goals generated by our engine involve gathering information. Not only is this a unique feature, but it's also one that could be used to model the dispersal of information, as well as surveillance, deception, and privacy. To facilitate this engine, we designed two AI systems: a quest AI to administer missions to players, and an agent AI to act as numerous placeholder players who are indistinguishable from human players. When a player performs an action such as entering a room or trading an item, our engine keeps track of that action and disperses the information to other players in the area. The quest AI may then use this information in the generation of missions. As our proof of concept, we created a game about espionage. Like a usual MMORPG, the characters maintain a variety of stats and skills, but rather than the usual qualities of strength and defense, our game tracks qualities like perception and stealth. We also have a number of factions -- groups with different goals. These factions may seek to spread, weaponize, safeguard, destroy, or profit from information.

Software Metrics Research

Problem Context: Requirement change analysis provides developers a means to maximize the productivity of a project. Without the analysis of software requirement implementation resources may be drained and the maintainability of a software system may be compromised. Aim: My aim was to compile and create a tool that is able to analyze software requirements and provide useful statistics and feedback to developers. Challenges: Issue trackers do not hold the code base that the issues are for. Querying for requirements and their effects needs at least two maybe even three separate entities of information. The requirement analysis tool needs the project to have an issue tracker and a code repository in order to be effective. Another challenge is the lack of information available from a code repository at any given revision. Git will not provide all the information necessary to perform a proper analysis of an entire project's requirements without rolling back to previous releases to obtain information. This is fine for smaller projects with lesser amounts of code. As the amount of code grows, so does the run time of the program due to it having to return to every single previous revision of a project. Results: I have created a beta release of a system that properly takes and analyzes the changes that happen to each class during the process of implementing a software requirement. Learned Topics: I learned about git and Jira APIs for python. I learned the importance of well-made requirements and the importance of proper issue tracking. I also learned a few things about databases as a result of working with large amounts of data. Next Steps: I will continue to work on this project as part of an Independent study CSC 400. The next steps are to improve upon the analysis tool and compute more metrics that are missing and necessary to provide a truly useful tool for developers. I am also writing a paper about the results of the projects which we plan to submit in a major internal peer-review conference such as the Mining Software Repository conference:

Reliability Evaluation of Silver Conductive Paste on TPU for Human Monitoring Systems

Wearable flexible electronics has the potential to improve the quality of health monitoring. Screen printing conductive paste for wearable applications has been explored; however, current capabilities of conductive pastes with pitches small enough for IC attachment is limited for wearable purposes. To characterize electrical reliability, DuPont PE873 stretchable silver conductive paste was screen-printed onto thermoplastic polyurethane (TPU) and underwent tests to determine how resistance changes after being stretched, washed, and dried. Resistance of circuit terminals was measured before, during, and after each test cycle. Failure was defined as a circuit’s resistance exceeding 100 ohms or the resistance increasing by 100%. In cyclic tensile tests, samples were pulled in tension for 100 cycles with no rest period. Samples passed several cycles before failure at 10% max strain at both strain rates, .008333/s and .04167/s. Samples would often fail early if pulled past 10% max strain. Resistance data showed a cyclic pattern with peaks at max strain. Both higher strain and higher strain rates led to larger increases in resistance peaks. During the rest period after 100 cycles, resistance measurements decreased over time, showing an elastic behavior. After about 30 minutes, the samples would return to a resistance of about 40-70% greater than the initial resistance. Thus, resistances above an acceptable range may return to an acceptable range given enough time. For washing and drying tests, printed samples were laminated onto stretchable polyester/lycra fabric. Some samples were encapsulated with DuPont PE773 to protect the printed traces while others had additional protection with silver epoxy covering bare terminals of encapsulated samples. Washing and drying was performed three times on the coolest and most gentle setting in a horizontal axis front-loading washer and tumble dryer. Bare samples failed after the first cycle. Encapsulated samples and samples with epoxy failed after three cycles. Washing and drying test samples also exhibited an elastic behavior with a resistance decrease of 5%-20% after 24 hours of rest. Both tests showed improved recovery and reduced plastic deformation in conductive pastes are needed for pitches small enough for IC’s in wearable applications.

Nitrogen management in aquariums using Bacterium consortium

  • Student(s): Jonathan Chen
  • Faculty: Nirupam Pal

The objective of this study was to understand feasibility of using nitrifying and denitrifying bacterium consortium in managing nitrogen problem in small and large aquariums. Ammonia is usually generated in aquariums due to dead material decomposing as well as fish food not being consumed completely by the fish. Ammonia and Nitrite are poisonous at low concentrations to fish and most healthy aquariums are kept at .1 ppm of both chemicals. Nitrate isn’t as toxic to fish and are usually kept at 15 ppm. Most aquarium filters at the moment filter out physical debris, and convert ammonia to nitrite and nitrite to nitrate to keep the fish healthy; however, algae use the remaining nitrate to grow causing a dirty fish tank. This results in having to remove the fish, clean the tank, and remove some water. The idea for this research was to create a bacteria medium filter that converts ammonia through the nitrogen cycle into nitrogen gas. A mixture of nitrifying and denitrifying bacterium mixture was used as the main bacteria source. Initially we developed a filter and growing bacteria on aquarium rocks. The system worked well on shaker scale (in vitro) but was not that successful in-vivo condition in a 10 Liters aquarium due to lack of carbon source, which naturally occurs in an old aquarium. Later, successful trials were performed with artificial carbon source. The student is currently working on the project during the fall quarter with aquarium water.

IoT Routing and Security Analysis

  • Student(s): Abineet Singh
  • Faculty: Bruce DeBruhl

In our Internet connected and data driven world the expansion of IoT (Internet of things) devices are skyrocketing. IoT devices like Amazon Alexa, Google Home, and Nest thermostats are extremely popular in today’s market. However, there is a much more diverse set of networked IoT devices including door locks, kitchen appliances and even consumer vehicles. IoT can provide automation and freedom to enhance our daily lives, but at a price. IoT devices can pose a major security and privacy risk, especially if security is not included by design. In October 2016, we saw the largest IoT cyberattack, where Dyn a major domain name service provider was victim to a distributed denial of service (DDoS) attack by a botnet named Mirai. While DDoS is not unique, Mirai was the first that used IoT devices including printers, baby monitors, CC cameras etc. This attack showed that IoT vulnerabilities need to be assessed especially since they are proliferating the market with an estimated 23.14 billion devices in 2018 and a staggering 75.44 billion in 2025. Over the summer I worked on building a customized router using an Intel NUC to read IoT traffic on a network. I built the router over the summer and, as a senior project, I am continuing the work on network configurations to classify traffic omitted by Amazon IoT devices. Continuing to work with Dr. DeBruhl, I look forward to designing classifiers, tests, and models for privacy/security risks of IoT devices using the custom router I built over the summer. Next steps will include understanding packet traffic, possible variances within power omission and application of network security protocols on the communication.

Modeling of Localized Transient Heating Within Fiber-Reinforced Polymer Composites Using FEM

  • Student(s): Dallas J. Johnson
  • Faculty: John C. Chen, PhD

The system modeled is a carbon fiber-reinforced polymer plate undergoing a vibration-induced heating process. The process is specific to a nondestructive testing method which detects the presence of delamination(s) through ultrasonic vibrations which generate localized heating at the delamination site(s) via viscoelastic and Coulombic damping within the sample. Several models were constructed using the commercial finite elemental software Abaqus. Convergence studies were conducted to establish modeling parameters for meshing, element selection, and time step size. Additionally, proper utilization of Abaqus interactions at lamina interfaces to model the thermal conductance present at delamination site(s) was examined. Qualitative benchmarking to existing experimental results was the primary method of calibration for the quantities of heat generation and thermal contact conductance; the results of the modeling indicate good agreement between the system and the model.

Software Requirements Use Within the CubeSat Community

  • Student(s): Noah Weitz and Luca Merlo Paula Soares
  • Faculty: Dr. John Bellardo

The aerospace industry is known for the amount of care it puts into properly designing and developing software requirements, since there are major consequences if a failure occurs. However, CubeSats do not fall under the same scrutiny. They lack requirements to ensure the final software product meets the same quality standards as the rest of the system. Therefore, it is common for problems to arise on CubeSats due to lack of software requirements. To help collect data and find possible solutions about this issue, research was done about software requirements methods used in the aerospace industry. Furthermore, it was also researched techniques used in previous studies to evaluate the effectiveness and usefulness of software requirements. Moreover, papers about problems that compare different software requirements methods were also saved for future reference. These papers will be used to guide the research and identify where the problems lie on and help solve them. Based on the research, some methods appear to solve some of the problems presented at the CubeSat industry. A modified version the Consortium Requirements Engineering behavioral model (CORE) for creating software requirements, for example, provides a usable structure for developing clear requirements for CubeSats. The CORE model was tested on ExoCube 2, one of the missions been developed at PolySat, showing improvement towards the team members’ understanding of the software requirements of the spacecraft. It was concluded that CORE is a good option as a model for creating software requirements for CubeSats; However, it require modifications to make it effective.

Effect of Multifunction Materials on Bio-Composite Structures Under Three-Point Bend Test

  • Student(s): Alexander Mandeville Bartlett
  • Faculty: Dr. Eltahry Elghandour

Traditionally, surfboards are pieces of Polyurethane foam wrapped in a fiberglass composite layer to add stiffness and strength to the member. Unfortunately the fiberglass composite process used is both hazardous to manufacture and difficult to reuse and recycle. This is where bio-composites can be applied, reducing the use of inorganic fibers through the introduction of natural fibers such as cornhusks, and hemp cloth. This project will present the manufacturing techniques and experimental test for integrated bio-composite beams under three-point bend test. The beams were constructed from multifunction materials (fiberglass, corn, and hemp) in five different combinations in order to discern the properties of each and come away with a plausible alternative to fiberglass. The structure consisted of a foam core and a 4 layer composite skin. Each composite variation had 4 samples to test (20 total) in order to gather the maximum amount of data and conduct a thorough analysis. The bend test was used in order to gather data focused on the stiffness and strength of each member. The samples with the cornhusks both held significantly less force than the others and actually broke into pieces when the load was applied. This left 3 viable samples; all hemp, hemp/fiberglass, and all fiberglass. The hemp behaved oddly as the composite structure never fractured resulting in a high specific strength at the cost of the materials stiffness. The fiberglass/hemp sample responded similarly to the all fiberglass, both had close ultimate strengths and stiffness’s. This concludes that surfboards could start using both hemp and fiberglass in the composite layup and the material properties would remain virtually identical while not adding weight to the board and at the same time significantly cutting down the amount of fiberglass.

Effects of Rice Derived Silica Aerogel in Insulating Hemp Concrete Composites

  • Student(s): Jose Urizar
  • Faculty: Eltahry Elghandour

A fundamental need within the residential and commercial construction industries for alternatives to petroleum and fiberglass based insulation has been identified. Major shortcomings identified with current insulations are: OSHA identified carcinogens, efficiency degradation over time, thermal bridges reducing efficiency, energy input to create fiberglass & reliance on petroleum. Potential answers to this fundamental need for healthy and cleaner forms of insulation are currently being explored. Proposed solutions include the shift towards entirely new materials more suited for sustainability all together which involves the scope of this project, to truly address the problem statement by developing a long-term solution that will not need to be revisited. Currently there is much research being conducted on insulating concretes which consist of bio-based aggregates. Cellulose-based insulators were a leading candidate for investigation due to its natural thermal insulating properties, its natural abundance and consumer familiarity, and ease of manufacturing. The investigation conducted took a closer look at high performing insulating concretes known as Hemp Concretes or more commonly Hempcrete. Hempcretes utilize naturally occurring binders and hurds, the inner core of the hemp plant which are inherently antibacterial and porous in nature. Rice derived silica aerogels were introduced to the hempcrete composites at differing volumetric fraction mix ratios to improve thermal resistivity. Silica aerogels are novel nanoparticles with high insulating properties which were manufactured through sustainable methods by GreenEarth Aerogels. The investigation conducted herein looked to incorporate the rice derived silica aerogels into the hempcrete, at varying mix ratios to investigate its effects on the overall insulating properties of the hempcrete composite created. Construction of a R Value measuring chamber and device was also investigated to approximate changes in thermal resistivity in accordance to ASTM C518. Results confirmed an increase in thermal resistivity as the amounts of silica aerogels were introduced to the hempcrete.

Building System Retrofits at Cal Poly- A Tool for Carbon Strategic Planning

  • Student(s): Jeff Wagner (ME) and Ben Christensen (ENVM)
  • Faculty: Dr. Gregg Fiegel

Cal Poly is developing a Carbon Action Plan to guide them to their goal of carbon neutrality by 2050. In support of this plan, Cal Poly Facilities is considering building retrofits related to existing heating ventilation and air conditioning (HVAC), lighting, water, and building envelope systems. Important goals of this project were: (1) to develop a methodology for characterizing the efficiency of existing building systems using available data; and (2) to forecast potential energy savings for different retrofit alternatives for building systems on campus. In support of the above goals, we developed a spreadsheet-based Building Energy System Project Tool (BESP Tool) to analyze and evaluate potential building retrofits on campus. Based on our research, we eliminated water system retrofits from our tool as they do not provide substantial energy savings potential. The BESP Tool incorporates the following building-specific data: floor area, lighting classification, HVAC system, envelope type, and desired retrofit levels. The tool computes reductions in lighting energy use and costs associated with the chosen retrofit based on per fixture assumptions. For the HVAC systems, the tool computes reductions in energy use and retrofit costs using a linear model developed from 2010 energy audit data. The tool's forecasting algorithms rely on data from energy modeling studies from a limited number of buildings (10 total) on campus. Therefore, the user should consider projections for energy and energy use intensity (EUI) approximate. The tool provides reasonable estimates of EUI for two 'blind' predictions. However, confidence in the tool will always be limited by the amount of empirical data used in its development and the fact that building HVAC, lighting, and envelope are complex systems that are difficult to model. Ultimately, the BESP tool provides Cal Poly Facilities with data to help inform carbon strategic planning efforts on campus.

p21 Gene Expression in Sphingomyelin Treated Keratinocytes

  • Student(s): Trevor Bingham
  • Faculty: Dr. Lily Hsu Laiho

According to the American Cancer Society, non-melanoma type skin cancer (NMSC) affects approximately three million people per year. NMSC impacts mostly keratinocytes, which are highly abundant within the epidermis. These cells are vital to skin barrier function, wound healing, and immune response. Though NMSC is highly treatable, there is a need to find novel strategies for protecting skin cells, such as keratinocytes, from UV photodamage. According to prior research done by this lab, sphingomyelin, a lipid found in the lipid bilayer of cells, appears to be an effective additive for reducing the quantity of UV photodamage in keratinocytes post UV-B exposure. To analyze this claim, p21 protein expression was observed. p21 protein functions in a host of manners ranging from DNA repair and cell cycle arrest to inhibition of pro-apoptotic proteins such as caspase-3, though overall p21 expression indicates cell damage. The goal of this research was to confirm prior claims utilizing different methods of analysis. To qualify cell damage, p21 gene expression was quantified by RT-qPCR in sphingomyelin-treated and untreated keratinocytes post UV-B. Cells were either incubated in 1% sphingomyelin solution or media for 24 hours and given 30mJ/〖cm〗^2 of UV-B. RNA was collected 24 hours after for processing and later RT-qPCR. For analysis, gene expression from each condition was normalized to untreated keratinocytes without UV exposure and each sample was normalized to GAPDH expression. Results from RT-qPCR showed a 23% decrease in p21 gene fold expression in sphingomyelin-treated keratinocytes relative to untreated keratinocytes, though this result was not statistically significant (2-sample t-test at 95% confidence). In tandem with prior results, it does appear that sphingomyelin does interact with the cells to reduce UV photodamage, yet a more extensive gene study must be done on cells with added sphingomyelin to assess its efficacy for skin cancer prevention.

Carbon Sequestration Using Magnesium Oxide Nanoparticles

  • Student(s): Shalto Dascher
  • Faculty: Professor Jean Lee

As global climate change worsens, investigation of new ways of effectively capturing carbon dioxide (CO2) is becoming increasingly important. This research examined the ability of magnesium oxide (MgO) nanoparticles that are simply and rapidly produced using a microwave oven to sequester CO2. Three routes of exposing the nanoparticles to CO2 were studied: (1) Injection of CO2 gas into the microwave oven during nanoparticle synthesis, (2) placement and sublimation of dry ice pellets into the microwave oven during nanoparticle synthesis, and (3) placement and sublimation of dry ice pellets in an airtight atomic force microscope (AFM) vibration isolation enclosure with microwave-synthesized MgO nanoparticles. In each of these three routes, AFM was the primary method used to examine whether CO2 was adsorbed by the nanoparticles. The first two routes of exposing the nanoparticles to CO2 did not appear to be effective in adsorbing CO2 onto the surface of the nanoparticles. Further examination of these samples using transmission electron microscopy (TEM) did not indicate the presence of any significant amount of carbon in the nanoparticles, suggesting that CO2 is not taken up by nanoparticles produced by either of these two routes. AFM examination of the third route of exposing the nanoparticles to CO2 revealed significant swelling of the nanoparticles as the dry ice pellets sublimed in the AFM enclosure. When the enclosure was opened to release the CO2 gas resulting from the sublimed dry ice pellets, a decline in the swelling of the MgO nanoparticles was observed such that the nanoparticles appeared to return to their size prior to CO2 exposure. These preliminary results indicate that microwave-synthesized MgO nanoparticles hold promise as a reversible means of carbon sequestration.

Endothelial and Smooth Muscle Cell Deposition and Morphology in Blood Vessel Mimics

  • Student(s): Benjamin Dennis
  • Faculty: Dr. Kristen Cardinal

This summer, my primary goal was to refine the protocols related to the Tissue Engineering lab’s Blood Vessel Mimic (BVM) creation and to establish consistent approaches for BVM development and cultivation. Specifically, I focused on evaluating the deposition of endothelial and smooth muscle cells within the BVMs (e.g. which cells end up going where?) and implemented methods to better visualize the resulting cell morphology. Before any of these analysis methods could be developed, I spent multiple weeks refining our cell culture protocols and evaluating the longevity and viability of cells from different vendors. After this study and a cost matrix analysis, I decided to proceed this project using the vendor Lonza’s Human Umbilical Vein Endothelial Cells and Human Umbilical Artery Smooth Muscle Cells. The process of evaluating the deposition of cells was executed after I set up the exhaust for our new critical point drier (CPD). From here I was able to refine the protocol to allow our PLGA scaffolds to be able to withstand the harsh environment of the CPD. Once sample BVMs were critically point dried, they were then sputter coated with gold and imaged on a scanning electron microscope (SEM). Furthermore, another method I used to visualize deposition of cells was through the use and development of a single CellTracker dye with a nuclear BBI stain*. Specifically, this allowed me to identify one cell type from another while in co-culture and hopefully on BVMs on future setups. During these projects I was able to do multiple vessel set ups for implementation of the SEM and Cell Tracker protocols described. Upon completion of this work, I am proficient in all of the core techniques for creating tissue engineered blood vessels and I was successful at investigating two new key assessment methods for evaluating cell deposition and morphology within our labs BVMs. *Poster presentation at the Surfaces in Biomaterials annual BioInterface Conference, in San Diego, CA

Carbon Sequestration Using Magnesium Oxide Nanoparticles

  • Student(s): Shalto Dascher
  • Faculty: Professor Jean Lee

As global climate change worsens, investigation of new ways of effectively capturing carbon dioxide (CO2) is becoming increasingly important. This research examined the ability of magnesium oxide (MgO) nanoparticles that are simply and rapidly produced using a microwave oven to sequester CO2. Three routes of exposing the nanoparticles to CO2 were studied: (1) Injection of CO2 gas into the microwave oven during nanoparticle synthesis, (2) placement and sublimation of dry ice pellets into the microwave oven during nanoparticle synthesis, and (3) placement and sublimation of dry ice pellets in an airtight atomic force microscope (AFM) vibration isolation enclosure with microwave-synthesized MgO nanoparticles. In each of these three routes, AFM was the primary method used to examine whether CO2 was adsorbed by the nanoparticles. The first two routes of exposing the nanoparticles to CO2 did not appear to be effective in adsorbing CO2 onto the surface of the nanoparticles. Further examination of these samples using transmission electron microscopy (TEM) did not indicate the presence of any significant amount of carbon in the nanoparticles, suggesting that CO2 is not taken up by nanoparticles produced by either of these two routes. AFM examination of the third route of exposing the nanoparticles to CO2 revealed significant swelling of the nanoparticles as the dry ice pellets sublimed in the AFM enclosure. When the enclosure was opened to release the CO2 gas resulting from the sublimed dry ice pellets, a decline in the swelling of the MgO nanoparticles was observed such that the nanoparticles appeared to return to their size prior to CO2 exposure. These preliminary results indicate that microwave-synthesized MgO nanoparticles hold promise as a reversible means of carbon sequestration.

Factor Augmented Reality (AR) Potentials into Facilities Planning and Design Research Project

  • Student(s): Dante Mazzanti, Andrew Allen, Tony Tran, and Jordan Moy
  • Faculty: Dr. Tao Yang

Our goal was to establish the elements needed to create an industrial engineering application relevant to facilities design and planning that we could incorporate into the Cal Poly IME curriculum. Very few IME programs in the country have virtual/augmented/mixed reality (VR/AR/MR) courses which represents a significant opportunity for Cal Poly to become a leader on this front. Through investigation of different head-mounted displays (HMD’s) we determined that the best options were the one that allowed virtual reality to be overlaid on the user’s surroundings. We called these optical see-through types of HMD’s. Specifically, we believe that our best candidates are the Microsoft HoloLens ($3,000), Meta ($1,000), or Lumus Display. On the software side the Unity VR Engine would be the easiest method of incorporating VR into the Cal Poly IME curriculum. This is because it uses Python to develop applications which fits in the IME department change from the VB programming requirement to Python in Winter 2017. The next step is to investigate how to incorporate elements from programs such as SolidWorks in a way that allows them to “snapped” to certain elements in reality (e.g. how to place VR objects such as workstation designs in a facility so that they are locked in place as the user moves around). We believe this technology is still nascent and offers a key area for Cal Poly to emerge from as a leader.

EPIC Digital Forensics Lab

The increasing growth and severity of attacks, combined with a relatively flat production of degrees in computer security, has led to a dearth of qualified security professionals. In particular, attracting a broad and diverse body of students to digital forensics has been challenging. The perception that digital forensics is the exclusive domain of white asocial males, that the field is devoid of creativity or individual expression, and the clear lack of positive, social impact have all been identified as root causes for disinterest and attrition. Thus, finding immersive, scenario-based educational materials is critical to addressing the national need for security education and exposing students to security as a potential career. The Engineering Possibilities in College (EPIC) summer program provides high school students from a variety of backgrounds such an opportunity. Our specific contributions created an experimental exercise of network traffic forensics generated from a real life wireless connection using technologies relevant to this audience and combined with a story-driven narrative to capture students interest. By capturing the network traffic on a wireless connection, the students access data about a “target” and their internet usage, demonstrating the major insecurities involved with open networks. Our preliminary results, as empirically measured through observation and informal student interviews, show that nearly all students found the exercise to be fun, engaging, and able to create new insights between technologies students currently use casually, and potential career paths in digital forensics. This early work lays the foundation for more extensive evaluation, and the further development of similar curricular modules useful in a variety of academic and social settings. We have packaged our curricular materials, and will be releasing them to the public on the website of our umbrella project,

Identifying PERILS of mining Github data and JIRA data

Problem Context: Github and JIRA projects have become important sources for software engineering researches. However, there are potential perils of mining data from them. My aim was to define and check a set of perils to help researchers measure the validity of the projects’ data. Challenges: API Rate Limits: My project exceeded the default rate limit of Github API after measuring one project. I have increased the rate by authenticating myself using personal access tokens on Github. Different names on JIRA and Github: Many projects use different names on JIRA and Github. I have resolved it by fetching the URLs of Github and JIRA from a JSON that contains metadata about all Apache projects. Limitations of APIs: GitPython does not provide users with detail about pull requests. I created a library that makes HTTP requests to Github to mine the data I needed. Results: I have successfully produced metrics of seven perils which show a low validity of data in both Github and JIRA projects. For example, some developers did not follow the rule of one commit per JIRA ticket. Learned Topics: I have learned three new APIs, GitPython, GitHub API, and JIRA API. Since the project involves data processing and parsing Strings, I have learned using regular expressions to speed up data fetching. Also, I gained more experience in object-oriented design. Next Steps: I will continue to work on this project as part of an Independent study CSC 400. The next steps are to measure perils for all the Apache projects and to develop a website application that allows other people to access our results. I am also writing a paper about the results of the projects which we plan to submit in a major internal peer-review conference such as the Mining Software Repository conference:

Computer Vision Strawberry Data Acquisition

  • Student(s): Jonathan Sato
  • Faculty: Jane Zhang

The goal of this research project was to collect numerical data from strawberry plants via taking pictures of them. Through different processes such as thresholding and morphological operations, the plant was able to be isolated from its background and measurements such as area and crown diameter were found. To convert the measurements from pixels to centimeters, a coin was used as a reference object to scale the measurements accordingly. The circular Hough transform was used to find the coin in the image. A Viola Jones object classifier was used to identify and count the number of flowers on each plant. This was accomplished by taking images of about 160 plants daily to train the classifier to find the flowers. All these different functions were then combined to output a spreadsheet containing numerical data based on a series of plant images. Further research can be done by creating a model to predict fruit yield using data obtained from the images. The collection of these images can be made faster and easier by using a drone to take the images of the plants.