Recognizing that midwives play an integral role in many small, decentralized communities within the developing world, the Portable Ultrasound Group aims to make it easier for them to diagnose the three most common antenatal complications: placenta previa, breech presentation, and multiple births. Obstetric ultrasonography is a highly effective method of diagnosing these conditions; however the cost is prohibitive in most developing countries. Additionally, ultrasonographic machines are often difficult to use and require more time to learn properly than is feasible for most midwives. Our aim is to reduce maternal mortality rates in the developing world by producing a substantially less expensive, easier-to-use ultrasonographic device which can be used by midwives to detect basic antenatal complications. We propose a device consisting of an ultrasonic transducer and a portable computer, connected VIA a universal serial bus (USB). The sonogram of the fetus will be displayed on the portable computer, allowing midwives the capability of analyzing these sonograms to diagnose basic antenatal complications and share these images with expectant mothers. Midwives will then be capable of transmitting these sonograms VIA a mobile phone network for further analysis and diagnosis at a nearby equipped healthcare facility. Potential complications include: Adequate and consistent power supply for the portable computer; Enhancing usability and the user experience for users (midwives) of the device; Avoiding an ethnocentric design. We will be discussing the process of designing the ultrasound device, as well as giving a brief demonstration of our current prototype.

Electronic Medical Records (EMRs) are becoming more commonplace in clinics and physicians’ offices, potentially changing the way physicians and patients interact. Consequently, it is essential to understand more fully the benefits and impediments of using EMRs in patient-physician communication. Previous research has already indicated that EMRs affect nonverbal communication (i.e., eye contact, body position) in the patient-physician communication and that nonverbal behaviors play an important role in patient-physician communication. Additionally, these nonverbal behaviors can have an effect on patient adherence to treatment. This study will synthesize previous findings and will examine the ways in which physician nonverbal communication interacts with the set up (i.e., spatial arrangement) of EMRs and how that interaction affects treatment adherence. Data will be collected via two questionnaires based on previous studies performed by Richmond et al. (2001) and Morisky et al. (1986). One questionnaire will assess the patient’s perception of his or her physician’s nonverbal immediacy behaviors. The second will measure the level of patient adherence to treatment. The spatial arrangements of the consultation room will be selected based on configurations identified by Mcgrath et al. (2007): open, closed, and blocked. There will be two physicians for each arrangement. The results of this study will potentially assist physicians in providing more efficient and quality primary care.

Past work with multiple input devices in low resource educational environments has largely been motivated by patterns of dominance that emerge when a large group of students is asked to share a single computer — inevitably, some students, at the cost of others, derive higher value out of the experience. Conversely, concern has been expressed about the lack of social learning and collaboration that results when students are given computers for individual usage. One means of addressing both concerns is to design multiple-input-based educational experiences to encourage collaboration among students. To explore this approach, we compared several different educational game setups in a controlled study with 192 primary school students in Bangalore, India. In the course of this study, we found that as collaboration was encouraged or enforced, patterns of dominance began to re-emerge, even though students each had direct interactivity with the computer. In this presentation, we examine and discuss these patterns as they relate to intervention design and environment.

The use of mobile phones in developing countries provides a practical way for economically, socially and governmentally marginalized people to locate the most profitable market to sell their goods. In areas where infrastructure is limited, transportation is a key problem. Not only does the choice of a distant market link to extra transport cost but the quality of goods are frequently degraded be the time the destination is reached. For example, the Maasai of Kenya drive their cattle on foot to market; the greater the distance, the greater the likelihood of increased mortality rates and general decline in health on the journey. Lack of knowledge from lack of effective ICT (information communication technology) adds to the challenge of efficient market choices. Mobile phones are often the only effective technology available. Published research into mobile phone use among marginalized people in the developing world is problematic. For one, most studies address issues that are unrelated to each other. Second, they rarely focus on marginalized, low-income, rural communities. This thesis contains a representative group of those often left out of mobile phone research. The Maasai, located in the remote Rift Valley are the subjects of my original research conducted from 2006-2010. Additionally, a qualitative instrumental case study (2007) provides the supportive framework, focusing on the uses of mobile phones in a commercial fishing industry in Kerala, India. Furthermore, an extensive quantitative study by Nairobi University (2005) supplies data for cell phone use in eleven sub-Saharan African countries. The results of these studies and the Maasai findings coalesce to form my conclusions that hopefully apply to similar communities across the globe. This thesis presents evidence that economically-marginalized people engaged in commerce can increase income levels by increasing profits through improved market information when mobile phones are available as a communication tool.

Does trust have a real economic impact? Intuition tells us that it does, but empirical evidence is difficult to find. The paper I am assisting with, Trust and Credit by Jefferson Duarte, Stephan Siegel, and Lance Young of the Foster School of Business seeks to measure the effects of perceived trustworthiness on loans made on the peer-to-peer lending website Prosper.com. From Prosper we have a real debt market in which users can create profiles, request loans, and fund loans. Requesters post credit information such as scores and histories, and may also include demographic information, education and employment history, and even photographs of themselves. In this environment the lenders presumably have all the information they need about each loan to reverse bid with interest rates on loans. My contribution is a web survey application I built to separate these photos from the loan postings and have the perceived trustworthiness of requesters evaluated on Amazon’s Mechanical Turk. Through the application we are able to crowd-source the survey to an "online marketplace" for an "on-demand, scalable workforce" where we had 3,284 unique workers evaluate 6,821 listings for trustworthiness, of which 3,291 listings were funded. By having each photograph evaluated by 25 different workers we are able to establish a consistent proxy for trustworthiness which is directly related to a loan listing outcome. With this proxy we find that a trustworthy person can promise an interest rate 182 basis points per annum lower than an untrustworthy but otherwise identical person.

The OpenJAUS is a software development kit for implementing unmanned systems, which uses Joint Architecture for Unmanned Systems (JAUS). JAUS was originally initiated by the Department of Defense to develop an open architecture for unmanned system and is now maintained by the unmanned technical committee of the Society of Automotive Engineers. Although the OpenJAUS is very useful for unmanned system practitioners, it is very difficult for non-experienced practitioners to use the development kit. For example, some OpenJAUS tutorials were missing detailed information of installing and running example applications. Only conceptual level diagrams of the architecture were repeatedly used throughout multiple OpenJAUS related research papers. Many parts of its Application Programming Interfaces (APIs) are not accessible through the Internet. No graphical version of a demo system, Vehicle Simulator, is available. In this research, by applying a reverse software engineering approach to the OpenJAUS, we produce several artifacts: visual models, Web based OpenJAUS API documents, and a graphical version of the vehicle simulator. As the reverse software engineering approach, 5W1H based re-documentation technique is employed to generate visual models in an industry standard modeling language that support multiple, concurrent, and hierarchical views to practitioners. From the models, the Web based APIs are generated and published onto a web site for the public access. In addition to the graphical demo system, these documents bring easier and more precise comprehension of the OpenJAUS to the unmanned system practitioners.