University of Vermont


SEED Projects 2007-08

Click the links below to read about the intriguing SEED projects our students worked on during 07-08.

Burlington Department of Public Works

"P5": Planning, Predicting and Plotting Plow Progress

Students: Yago Alonso-Garcia (EE); Matt Clark and Alex Estey (ME); Tristan Schneitter and Ken Austin (CS)

In winter storms, the City of Burlington Department of Public Works (BDPW) clears approximately 120 miles of two-lane roads (240 lane-miles). Snow fighting operational plans are designed to adapt to various inclement conditions defined by storm intensity (inches per hour), forecasted duration, air temperature, and the availability of snow fighting resources (staff and equipment). Currently supervisors check roadways and the progress of individual operators via two way radios. Calls are received from the public and emergency services for service and a complaint list is created for supervisors to assign work to individual operators. Various accounts of what occurred during a snow event often conflict with each other. With ten plows, supplemental trucks as needed, three baseline snow fighting plans, 14 full time employees and additional staff when necessary, clearing the roads while keeping track of all relevant information is a difficult task.

Clear roads and quick information
Students will create a system to collect, process and display real-time GPS-based information on the location and route history of snow removal and de-icing vehicles deployed by BDPW during and after storms. The goal is to maintain up-to-the-moment central office and citizen awareness of the recent history of plow location in order to alter plans as required by changing conditions, as well as predict snow clearing schedules for as-yet uncleared roads, and respond quickly to citizen concerns. The data must be secure and accessible interactively from any web-enabled computer and displayed so as to be easily queried and interpreted by a person of entry-level computer familiarity.

The contact engineer is Norm Baldwin and the faculty mentor is Jeff Frolik. For more information, see Burlington Department of Public Works.

Return to Top

Central Vermont Public Services

Making it Easier to Use Cow Power

Students: Geovanny Rodriguez (EE); Asa Parker and Jase Skellie (ME)

The students will design, fabricate, test and deliver a working prototype of Cow Power "MOM" — Monitoring and Operation Module. The Central Vermont Public Services (CVPS) Cow Power project is a well-established experiment in generating electric power on working farms by anaerobic digestion of manure. According to the CVPS web site, "To generate the biogas fuel, manure is held in a sealed concrete tank at the same temperature as cow stomachs, 101 degrees. Bacteria digest the volatile components of the manure, creating biogas while killing pathogens and weed seeds. The biogas, which is part methane, fuels an engine/generator, and the energy is put onto CVPS's power lines for delivery to customers."

MOM will watch for the system
MOM will be a system of sensors, communication links, data processing, control lines, actuators and farmer interface for a current experimental Cow Power installation. Its purpose will be to capture records of significant system variables, generate control inputs based on current and past values of these variables, and provide the user with a display screen exhibiting status, trends, context-dependent instructions, alert flags and other useful and appealing graphical, numerical and text information. A major design goal will be usability to non-technical users. Another will be some degree of software-controlled automation. This work will be driven by thorough understanding of the users and the circumstances of use — as well as by technical input from the system developers at CVPS.

Two major systems, "engine control" and "generator control," are currently separate for farmers. Engines have a variety of sensors such as engine oil temperature, pressure, coolant temperature, exhaust temperature, and so on. Generator sensors include current, voltage, frequency, and electric relay status. "Adding digester pressure (inches of water) and gas flow and gas components, btu meters on engine heat loops and digester heat loops, differential calculations, and displaying results on one interface would add value to the system," says Josh Bongard, faculty mentor, "as would adjusting generator output with direct feedback from gas pressure."

CVPS contacts are: David Dunn, Gregg White, Rob Nelson, and the Rowell family at Green Mountain Dairy. The faculty mentor is Josh Bongard. For more information, see:

Return to Top

DR Power

Z-turn Lawn Mower

Students: Alyssa Kennett (EE); Chris DiLallo, Chris Kaszas, Kerry Klauder and Nick Sacco (ME)

Working with the engineering staff at DR Power, this student team will develop a prototype or test bed for a new "zero-turn" push/ride mower which will run on electric power. The mower will represent a convergence of the industrial design look and feel of the Neuton light-duty electric product line and the beefier DR Z-turn lawn mowers which are based on internal combustion engine-driven hydraulic pump/control valve/motor systems. Although these systems have proven to be durable and safe, they are not optimal when energy conservation, low noise and minimized air pollution are given greater weight in the design. The target user of the new product will be a 45-year-old rural resident with a 2-acre lawn who is looking for an alternative to gasoline, spark plugs, filters, oil changes, noise and fumes.

Making the charge last
To successfully scale up the cordless Neuton technology, attention must be directed to the new design's energy budget and to reducing energy waste in order to make the current-technology rechargeable battery packs last longer. One design focus may be advanced blade design and improved air flow under the housing in order to cut with greater energy efficiency. A potentially useful technology may be hub motors, and possibly power-assisted manual drive (patterned after wheelchairs that use this approach). Depending on scope and priorities, attention may devoted to aesthetics and human factors, even on the first prototype.

The contact engineer is Jim Peterson and the faculty mentor is Robert Jenkins. For more information, see:

Return to Top

General Dynamics Armament and Technical Products

What's the Problem? Collecting Data on the Assembly Line

Students: Charles Veronneau (EE); Ryan Grimaldi (ME); Heather Truax (ME and Math)

Manual machining and assembly operations have limited automation. This means that collecting data on glitches cannot rely on built-in control and sensing channels. Currently, information at problem locations — at General Dynamics and elsewhere — is collected by a supervisor or an industrial engineer assigned the task. This SEED team is working to create a low-cost user-configurable instrument package to meet this need called, "A Supervisory Control and Data Acquisition (SCADA)." SCADA will be used for shop-floor-level collection of process status and milestone data. It will collect data from sensors and machines (on shop floors or in remote locations) and send information to a central computer for management and control.

Tough requirements
The new SCADA interface must be easy to use and durable enough to meet the tough shop floor requirements, including, in some settings, resistance to explosions. In addition, data transfer (e.g., Ethernet, RF, etc.) must work in all ATP shop environments. Also, the data processing software must be simple and user friendly with little maintenance required.

GDATP contact engineer is Frank Uhelsky. The faculty mentor is Mike Rosen. Visit these websites for more background information:

Return to Top

Green Mountain Coffee Roasters

Saving the Environment, One Cup at a Time

Students: Matt Seekins (EE); Zach Burchman, Travis Gang, Ryan Neary and Dana Geer (ME)

Use of the new Keurig® Single-Cup brewing systems — see the one in the School of Engineering office, for example — is in excess of one million cups a day and production rates are increasing. The SEED K-cup team members are working to develop a device that can disassemble finished cups so they may be sorted for either composting or recycling. Green Mountain Coffee Roasters (GMCR) plans to expand the application for individual office and home use, but the immediate use will be for test sample cups and rejects from the manufacturing floor in Waterbury. If successful, 50,000-80,000 cups per day with four packaging elements (coffee, plastic cup, filter paper, and aluminum foil lid) will be separated into homogeneous streams of potentially reusable materials.

Not so easy
Because heat is used to bond the various elements to prevent oxygen, light and moisture from degrading the coffee and spoiling the product, it is difficult to separate the K-Cup after use into its individual components. Without the seal, the coffee quality and freshness is compromised. "As Keurig® Single-Cup brewing systems grow in popularity, the environmental impact of the K-Cup® waste stream is one of the most significant environmental challenges we face," says Paul Comey, Vice President for Environmental Affairs at GMCR.

In addition to Paul, the GMCR SEED team's contact are Jason King and Wade Hodge. The UVM faculty mentor is Mike Rosen. For more background, visit these web sites:

Return to Top


Lithography Productivity Challenges

On the IBM fabrication line in Essex, 200mm-diameter wafers of chips progress through a sequence of photolithography processes in boxes holding up to 25. Each wafer is repeatedly passed through three steps — spin casting with a photoresist, optical etching, and mechanized transport. A design challenge is embedded in each of these steps and a SEED team has been assigned to each.

* * *

Spin Casting: Don't rain on my wafer

Students: Greg Burtt, Greg Hewitt and Dave Valente (ME)

Spin casting is achieved by spinning the wafer at speeds up to 6000 rpm with stepping motors while a drop of photoresist is applied to the center of the wafer. The photoresist spreads across the wafer, reaching its final thickness with a high precision. Most of the applied photoresist is spun off the wafer and is collected in a "coater bowl." There is a controlled airflow around the wafer edge and a solvent rinse of the wafer edge and back side. This process has been encountering problems. Aerosols are sometimes created in the bowl which backstream up and "rain" on the wafer. Bowl surface defects and vortex effects between the bowl edge and the wafer edge may both contribute to backstreaming of atomized material. This SEED team will develop a test bed to more fully characterize the phenomenon, and specify the necessary design changes.

The contact engineer is Kevin Remillard and the faculty mentor is Jeff Marshall.

* * *

Optical Etching: Do you know where your reticle is?

Students: Andrew Chamberlin, Emir Heco, Nathan Roumanis and Brad Williams (ME)

Circuits are etched on a photoresist that has been cast on the wafer repeatedly, in successive layers. The optical object that forms the image is a plate of glass with the necessary pattern etched on it with an accuracy of tens of nanometers. There are about 4000 of these reticles, each stored in its own protective clamshell case, in a storage area adjacent to the fabrication line. In sync with a production schedule, each needed reticle is obtained from storage, transferred from its clamshell to a standardized pod, and hand-carried to an entry port on the etching "tool" (read "incredibly precise complex electromechanical software-controlled machine"). At this point, a bottleneck can develop. 1600 wafers may be processed each day, in full production. Despite the fact that, during a particular etching process, each wafer may need up to six reticles, there are only two entry ports for the reticle pods. At a particular point in time, 250 reticles may be loaded in their pods and in repeated sequential use. In short, a new mechanical and human operator process is needed to improve the availability of reticles when and where they are needed. This will be the challenge for the SEED team.

The contact engineer is Paul Sonntag and the faculty mentor is Mike Rosen.

* * *

Robotic Transport: More precision, less debris

Students: Andrew Vize (EE); Matt Dulude, Greg Hewitt and Mike Johnson (ME)

Robot for handling digital chip wafers

Robot for handling digital chip wafers during photolithography at IBM.

As a wafer travels through the sequence of spin casting and etching steps, a special-purpose robot moves it from the load/unload port to the process chambers and back. These robots and their plastic-coated, stainless steel drive cables were designed and purchased over twenty years ago. Their age is showing. When a robot needs maintenance, which is typically about once a year, a rebuild requires 15 hours. Swapping in a rebuilt unit for one about to undergo maintenance takes 12 hours. In terms of productivity, this is very expensive.

The signs that a rebuild is needed are not subtle. Wafers start to show signs of foreign material generated by the robot or — more obviously — wafers start to break. This SEED team will develop modifications for these robots to deal with stretching of the drive cable, backlash in a gear train, a slide mechanism that generates fine debris, and other problems. They will need to live with the constraint that any redesign of the robot change must work with existing "locked" software.

The contact engineer is Dan Forcier and the faculty mentor is Mike Coleman.

* * *

For more information on these IBM projects, see:

Return to Top

Local Motion

Bikes across the water

Students: Tom Kidd (EE); Sean Horning, Marco Maffeo-Robinson, Justin White and Ross Ziskind (ME)

Local Motion is a non-profit Burlington business promoting bicycling, providing services to bicycle users, and encouraging alternative transportation facilities in Northwestern Vermont. For a brief period each summer, Local Motion runs a ferry to carry bikers across the Colchester/South Hero causeway for access to the Champlain Islands. The ferry service (supported by local municipalities) wants to offer daily seasonal service. Their vessel, a 30-foot pontoon boat with a 115 hp gasoline-fueled outboard, currently logs about 3500 boardings. In keeping with Local Motion's emphasis on green lifestyles, the goal of this SEED project is to reduce or eliminate dependence on fossil fuel for transportation of bicyclists across the cut.

Numerous possibilities
The conceptual phase of the project will identify, analyze and rate alternative power sources and their combinations. These may include biodiesel, stored and real-time human power, accumulated and instantaneous wind and water power, solar power, and their practical combinations. More generally, the team will consider alternatives to the current pontoon boat: cable ferries, moveable bridges, submersible bridges, an aerial tram — all possibilities are on the table. The deliverable will be a detailed design for a greener means of crossing the cut and a working prototype of the system or critical/innovative subsystems. The project will include on-site testing (in the Spring, weather permitting) as well as bench simulations. Conceptual design decisions will be generalizable but constrained by local data on the wind and sun at the cut; expected patterns of use; regional economic considerations; and attitudes of stakeholders including the state legislature, Local Motion, and ferry users.

The Local Motion contact is Island Line Project Coordinator Brian Costello. The faculty mentor is Mike Coleman. For more information, see Local Motion: Colchester / South Hero Causeway.

Return to Top

NASA Exploration Systems Mission Directorate (ESMD)

(via the NASA/Vermont Space Grant Consortium [VSGC] funded by the ESMD's 2006 Higher Education Funding Competition)

Design & Control of Miniaturized Positioning Thrusters

"Nanosats" — miniaturized orbiting spacecraft — show great promise for the next generation of NASA missions. Having masses of 20 kg or less, these spacecraft will be capable of operating in distributed networks ("formation flying") and performing mission objectives not currently achievable with traditional satellite architectures. Their propulsion systems must be capable of providing extremely low levels of thrust and impulse for orbital maneuvering and precise station keeping while also satisfying stringent demands on size, mass, power consumption and cost. To realize this vision, NASA must explore and evaluate innovative engineering approaches to the design and development of miniaturized propulsion systems for nanosats.

Benchtop test bed

Proposed test bed for mini-thrusters

Schematic of proposed test bed for mini-thrusters to be used by NASA on "nanosatellites."

This SEED team is working to design, build and test a miniaturized "orbital positioning system" capable of keeping a virtual "satellite" aligned with a moving target source. For simplicity, movement will be restricted to one or two degrees of freedom of translation. The prototype system will require design of the miniaturized propellant delivery subsystem, thruster nozzles, on-board sensors for target tracking, and a closed-loop control system. For simplicity, a cold gas propulsion system will be used in the first prototype. The plan is for a subsequent prototype to feature a hydrogen-peroxide-based monopropellant scheme, with an appropriate catalytic chamber.

Research collaborators at the NASA/GSFC Propulsion Branch will participate as external advisors. The faculty mentor is Darren Hitt. For more information, see:

Return to Top

National Science Foundation

Earthquake Engineering Simulation Research Program: Measurement of the Strength of Liquefied Soil in Centrifuge Models

Students: Michael Parks (EE); Kyle Bowley and Stefan Desis (ME)

The National Science Foundation (NSF) has invested over $75 million to build 15 earthquake research equipment sites throughout the U.S. The goal is to accelerate progress in earthquake engineering research and to improve the seismic design and performance of civil and mechanical infrastructure. One focus is on saturated sands which, subjected to earthquake loading, experience drastic loss of strength and behave as heavy fluids. As long as the liquefied state persists, the soil will flow down slopes, producing destructive landslides and large drag forces on obstacles such as foundations. Modeling this behavior for risk studies and engineering design requires research using a centrifuge-based test bed that simulates real-world situations. The efficacy of alternative designs or seismic retrofitting techniques can be compared in repeatable scientific tests. A soil model is placed in a container on a shake table that can be triggered while the centrifuge is spinning. One can instrument the model with a variety of instrumentation including accelerometers, miniature pore water pressure transducers, strain gages, displacement transducers and high-speed video to model the model response.

Shifting sands
The shear strength of liquefying sand can be measured in-flight (at the end of the spinning centrifuge arm) in a seismic centrifuge model using a thin plate, about 25 x 25 x 1.5 mm pulled horizontally through the sand sample, with its major dimensions parallel to the base of the model. By measuring the drag force on the coupon, it will be possible to observe the evolution of the soil shear strength as it decreases and subsequently increases as pore water pressures dissipate. In this SEED project, the students will develop the instrumented test chamber for PI Mandar Dewoolkar's research. Their work will require coordination and collaboration with technical staff and researchers from UC Davis and possibly a visit to their centrifuge facility. The student-designed unit will be built and tested in the lab at UVM. A civil engineering student will join the team to help with the soil model construction and testing.

The faculty mentor is Mandar Dewoolkar. For more information, see:

Return to Top


Outerwear test and demonstration system

Students: Eduardo Sanchez (visiting EE student); Edward Skiba and Matthew Tefft (ME)

Orage is a world leader in blending high-tech fabrics with fashion-inspired styling to create a highly popular line of ski outerwear. Orage jackets can look like denim, canvas, or suede, but their fabric technology system, known as PRIME, functions as a highly effective and comfortable barrier against the elements on the mountain. This SEED team is developing a system to measure and exhibit the water repellence and breathability of Orage garments. This is actually two projects in one. On the factory floor, a repeatable objective sensitive means is needed for characterizing and comparing garments in ways that relate directly to user comfort, insulation and protection. In shops, showrooms and tradeshows, a more showy system is needed to vividly demonstrate the advantages of Orage garments to a lay audience of potential customers.

Watch or feel?
One issue the SEED team is contending with is whether numerical and other visual readouts, a direct physical demonstration, or consumer-interactive subjective demonstration will be most effective — or some combination of these modes. Further, an objective quantifiable index of "breathability" may need to be developed and the system will almost certainly need to be readily transported and deployed in order to be useful for the tradeshow circuit.

The Orage contact is Jamie Brandon and the faculty mentor is Yves Dubief. For more information, see Orage.

Return to Top


Probecard Validation System

Students: Gregg Carpenter and Chris Hudson (EE); Tristan Schneitter (CS)


Probecard used at Qimonda in the testing of wafers of digital chips during manufacture.

This team will design an automated software-based system for visualizing and validating a "Probecard" prior to manufacture. This system must be intuitive, standardized and well-documented. A Probecard is a critical piece of equipment necessary for every test setup in semiconductor manufacturing. It carries the electrical signals between the test equipment and the wafer of silicon devices under test. A Probecard is uniquely designed for each product and costs $100,000 to manufacture.

Design will reduce error and save money
A mistake in the Probecard design process causes direct capital loss for redesign and lost revenue due to the product's delayed market introduction. The Probecard validation process is currently very manual and has few standardizations or automated checks. This makes the process error prone. An automated validation tool could significantly reduce the risk of design and ordering mistakes.

The Qimonda contact engineer is Barry Hulce. The faculty mentor is Tian Xia. For more information, visit Qimonda.

Return to Top

SnowMAN — Snow Measurement Analysis Network

(Funded by the NASA/Vermont Space Grant Consortium [VSGC])

Students: Matthew Casari and Christopher Henwood (EE); Jonathon Ewing (ME); Honors Thesis Collaborator: Charlie Robinson (CS)

Snow Water Equivalent (SWE) is the measurement of the water content held within snow. In recent years the importance of being able to accurately measure SWE has been recognized by people in the agricultural and drinking water industries. Seasonal snow melts are the determining factor behind aquifer volume and flow rates. The successful prediction of these variables allows users to prepare compensatory measures for drought or flooding conditions to come. Currently, the only systems that can measure SWE are large-scale permanent facilities that depend on hardened shelters and sophisticated transmission technologies. Their cost and staffing requirements make them inaccessible to many potential users. Further complications arise with the attempt to weigh snow. Current systems include a device known as a "snow-pillow," an instrumented bladder filled with anti-freeze. This system is fraught with problems, the most serious being inaccuracy when bridging occurs, for example, when the lowest layer of snow is no longer in contact with the ground due to the tunneling of animals, melting of snow, or snow-pack shifting.

Networked snow sampling
This SEED team, which began work in Summer 2007 because its members will graduate at the end of Fall term, is addressing two challenges: develop a system that is portable and power efficient, and uses wireless "Motes" to connect with others via a wireless network; and design/build a new method of weighing snow that can be utilized by the Mote network and deal effectively with snow-bridging. The total system will measure (at least) snow weight, snow height, temperature, and humidity. The system must also be able to withstand upwards of 150 days on a Mount Mansfield site presently used by University of Vermont Snow Hydrologists to make manual measurements of all these variables.

The faculty mentor is Jeff Frolik. For more information, see:

Return to Top