Recent Projects: Product Design

Eye Tracking Device

Team: Nathan Lehrer, Grant Neubauer, Tyler Ray, John Zahka
Sponsor: Analog Devices Inc.
Technical Liaison: Jim Doscher
Faculty Advisor: Brian Pogue

The way we interact with technology has changed as often as technology itself. The evolution of this interaction has included the keyboard and mouse, touchscreen, and even voice control. Eye tracking, a technology that originated as a research tool, has started to enter consumer markets as an alternative method of human-computer interface. One great potential use for eye tracking is in the emerging field of glasses-based computing, such as Google Glass.

Eye Tracking Device Analog Devices' new low power optical sensor, which measures the average position of an object in front of it, may offer a way around using cameras and power-hungry image recognition algorithms in eye-tracking devices. If the data from this sensor can be correlated to a user's point-of-gaze, this technology has the potential to successfully integrate portable eye-tracking and glasses-based computing.

We evaluated the capabilities of this sensor as a reliable eye tracker. The deliverables of this project were: (1) Create a viable design for an eye-tracking system; (2) Construct an experiment, test apparatus, and prototype to test viability of the design; (3) Document test results; and (4) Develop a demo to showcase the abilities of the device and compare with existing eye tracking systems.

The design incorporates an unprecedented method called "scleral averaging," which relies on the fact that the output of the sensor is related the amount of visible sclera, which changes when the eye moves. Experiments determined the ideal parameters for use of the sensor as an eye-tracker. An experimental procedure was designed and executed in which subjects were stabilized in a head fixture and instructed to look at a series of known points while sensor data was collected on their point-of-gaze. The sensor's performance was evaluated using a nonlinear regression algorithm that determined the error between the true position and the sensor's output. The testing showed that in ideal conditions, the sensor has the accuracy, precision, and range necessary to determine in which cardinal direction the user is looking, but not enough accuracy for higher resolution applications. Thus it could be useful for tracking "gestures" (i.e., a left-to-right or up-to-down eye swipe) or which box in a 3-by-3 grid the user is looking at.

First Responder Heat Exposure Monitoring Device

Team: Quinn Harper, Alexander Murphy, Elias Stein
Sponsor: Boston Engineering
Technical Liaison: Mark Smithers
Faculty Advisor: Stephen Taylor

First Responder Heat Exposure Monitoring Device Recent and significant improvements to turnout suits now allow firefighters to remain in several hundred-degree heat for longer periods of time. The improved insulation results in the gear being equally good at retaining the firefighter's body heat inside the suit as keeping out the external heat, resulting in a significant increase in injuries due to heat stress among firefighters. However, modern turnout gear does not give firefighters a way to monitor their risk of overexertion and heat illness. Our sponsor has endeavored to produce a prototype device that measures the temperature within a firefighter's turnout gear, and relays this information to a command post outside the fire zone. Our project was to research, develop, and produce a works-like prototype. The embedded software and finite state machine have been developed and are running on the prototype electronics. The enclosure has been redesigned to include a secure clip mechanism and simplified battery system. The prototype wirelessly transmits temperature data with ±0.5 degrees Celsius, while demonstrating an estimated battery life of up to 160 days.

Mobile Tackling Target

Team: Quinn Connell, Noah Glennon, Elliot Kastner, Andrew Smist
Sponsor: Friends Of Dartmouth Football
Technical Liaison: Lucas Hussey
Faculty Advisor: Minh Phan

Mobile Tackling Target Football teams need a dynamic and mobile device that simulates player motion as realistically as possible in order to practice safe tackling form without the injury risks associated with player-to-player contact. The solution must reflect the unpredictable motion of a live player, be safe to tackle, simulate realistic tackling, and work on a turf field in all playable weather conditions. Deliverables for this project are a full-scale working prototype, an extensive patent search and report on state-of-the-art, and a realistic business plan. We researched various solution strategies and verified that a mobile dummy would be the most effective solution for the problem. We designed a remote control dummy driven with a ball drive and built and tested a full-scale prototype.

Our sponsor and our group have filed a provisional patent application that covers all aspects of the device.

Personal Nicotine Exposure Monitoring Device

Team: Yves-Marie Duperval, Kwame Ohene-Adu, Tewende Sawadogo
Sponsor: Tuck-Fresh Air Sensor Corp.
Technical Liaison: Joseph BelBruno
Faculty Advisor: Kofi Odame

Personal Nicotine Exposure Monitoring Device Smoke that is exhaled and/or released from the end of a burning cigarette is defined as second-hand smoke, while the hazardous residue from smoking that clings to hair and fabric is called third-hand smoke. Second-hand and third-hand smoke exposure have been identified as carcinogens and contribute to increased risks of cardiovascular diseases and childhood ailments, according to a 2006 report published by the U.S. Department of Health and Human Services. Third-hand smoke can be especially harmful to infants.

Research conducted at Dartmouth over the last few years resulted in the development of a nicotine sensor, which uses a conductive polymer film that changes its resistance in response to the ambient concentration of nicotine. Our sponsor acquired the sensor and developed a device that interfaces the sensor with a processing unit and allows the real-time collection of the sensor's resistance. However, this solution is costly, bulky, and unable to properly implement a stable sensor connection.

Our group proposed an innovative solution that accurately reads the sensor's resistance and the ambient temperature in order to provide users with a quantitative assessment of nicotine in their surroundings. We produced a preliminary printed circuit board (PCB) that provided a platform to conduct tests. A second iteration of the PCB incorporated the sponsor's requirements. Extensive testing was then carried out to verify the device's compliance with the specifications. In addition to submitting a functioning prototype to the sponsor, we provided documentation for reproducing our design and a technical manual for operating the device.

USB-Plugged Data Logger

Team: Alex Hanson, Yihan Li, Curtis Lim
Sponsor: Linear Technology Corp.
Technical Liaison: Sam Nork
Faculty Advisor: Jason Stauth

Linear Technology Corporation is an analog circuits company that developed a high-performance 20-bit analog-to-digital converter (ADC). It can be used in data acquisition devices such as oscilloscopes and certain medical devices. Although they had a generic demonstration board to help market the ADC, Linear Technology was interested in seeing the device integrated into an impressive, easy-to-use USB-plugged data logger. The goal of the data logger was to demonstrate the capabilities of the ADC in a compressed, plug-and-play package. The project had hardware, firmware, and software components that required a breadth of expertise from the students and engineering contacts at Linear Technology. The hardware involved integrating the ADC with circuitry that acquired voltage inputs, interfaced with a computer through USB, and provided power to the device solely from USB power. The firmware was responsible for receiving, timing, and sending the digital bits from the ADC to the USB circuitry. The software, sourced in Visual Basic and C#, allowed the user to control the data logger; it had a voltage-over-time plot and table, a digital code count histogram, and data export functionality. The end goal was for the software to automatically recognize and operate the data logger.

USB-Plugged Data Logger