This paper explores the challenge of taking Global Goods software to international scale. Global Goods software is non-commercial software designed to support global development goals. We argue that a fundamental challenge behind this type of software is the different roles of the global organizations that fund projects, the country leadership that controls implementation, and the actual users of the software. To address this, it necessary to have a design process that balances interests of stakeholders and a technical design that allows for modularity and extensibility. We present a case study of an application for country level management of the immunization cold chain that we have developed and contrast it with major Global Goods software systems such as DHIS2 and OpenMRS. The contributions of the work include the design of a pipeline for building a Global Goods application that is deployed across multiple countries, a collection of lessons learned during system design and implementation, and a comparison of extensibility strategies of different global goods applications.

In resource-constrained communities, organizations often use information and communication technologies to amplify their limited resources to improve education, health, and economic opportunity. Over two-thirds of the world's population have mobile phones, yet less than half are connected to the Internet [23]. Organizations helping disadvantaged populations often rely on mobile devices as their primary computing resource because of their availability in resource-constrained contexts. However, to reach under-served populations, mobile applications often operate in areas with no connectivity or challenged network environments. Unfortunately, many mobile application frameworks are generally not well-suited for long periods of disconnected data collection and management. Furthermore, mobile application frameworks are generally aimed at users with significant technical skills and resources. In this paper, we discuss our experiences building, deploying, and refining the Open Data Kit (ODK) 2.0 tool suite. ODK 2.0 is a modular application framework that facilitates organizations with limited technical capacity to build application-specific information services for use in disconnected environments. We discuss ODK 2.0's flexible abstractions that enable users of varying technical skill levels to create customizable mobile data management solutions. We present ODK 2.0 case studies involving multiple organizations and discuss lessons learned from building a service-based mobile application framework for disconnected data management.

Open Data Kit (ODK) is an open-source, modular toolkit that enables organizations to build application-specific mobile information services in resource-constrained environments. Feedback from users and developers about limitations experienced with the ODK 1.x set of tools led to a redesign of the system architecture and the creation of new tools. This demonstration presents a revised tool suite called ODK 2.0. This expanded ODK toolkit aims to increase an organization's data collection and management capabilities by supporting data synchronization, adaptable workflows, more configurable presentation screens, and increasing the diversity of input types by enabling new data input methods on mobile devices.

Mobile devices are integral to the workflows of many organizations working in rural or disconnected contexts. Data is often collected on mobile phones and tablets using tools like Open Data Kit (ODK). However, some of the applications require users to revisit and update previously collected data, necessitating easy viewing of stored data. To make it easier for organizations to create flexible information services, we present ODK Tables, an Android tool that allows users to enter and curate data on mobile devices. Tables leverages web tools to make mobile app creation simple. It provides abstractions to make the process straightforward and allows app designers to access data through a JavaScript API. App designers can create a custom app using only a small number of HTML and JavaScript files. This facilitates the creation of a custom user interface but leaves storage, data management, and synchronization to the framework. The result is a fully featured Android app based on established web-based tools with full support for disconnected operation.

We describe how current radiological best practices are predicated on a sophisticated technological ecosystem usually comprised of multiple large-scale displays, and integrated record keeping and communication systems driven by high-speed networks. At the same time, current development of low-cost ultrasound (U/S) devices for low-resource settings trends towards small-scale, independent devices with palm-sized screens. We reviewed existing literature, analyzed findings from two years of fieldwork in Uganda, and conducted an interview study with clinicians about radiology work practices to determine which patterns and technologies contribute to the efficacy of ultrasound. We use these findings to inform how ultrasound technology in low-resource settings can most usefully be developed and deployed. In addition, findings are relevant for creating medical technologies for low-resource environments generally, as we make clear the importance of considering not just technology development aspects like power consumption and interface, but also larger technology and work ecosystems.

Ultrasound imaging is an effective tool for identifying maternal mortality risk factors. However, the high cost of ultrasound devices and the scarcity of ultrasound training are two major barriers to adoption in the developing world; complicated and expensive medical technologies are unlikely to meet the needs of users with limited opportunities for formal training and continuing education. To address these particular barriers, we have designed an inexpensive ultrasound system composed of off-the-shelf hardware and custom software. Our system is designed for use by rural midwives---often central medical figures in resource-constrained communities. This paper presents our work integrating help and tutorial software into the ultrasound system we developed. Our goal is to supplement a midwife's conceptual and operational knowledge of diagnostic ultrasound through appropriate and adaptive scaffolding. The design of our help system is informed by interviews with sonographers and radiologists in the United States and Uganda, and data collected during two fieldwork visits to Uganda. Through our fieldwork and preliminary evaluations, we have found that in addition to in-context reference material accessible during an exam, midwives need in-depth learning materials that can be accessed outside of a medical exam scenario.

Smartphones can now connect to a variety of external sensors over wired and wireless channels. However, ensuring proper device interaction can be burdensome, especially when a single application needs to integrate with a number of sensors using different communication channels and data formats. This paper presents a framework to simplify the interface between a variety of external sensors and consumer Android devices. The framework simplifies both application and driver development with abstractions that separate responsibilities between the user application, sensor framework, and device driver. These abstractions facilitate a componentized framework that allows developers to focus on writing minimal pieces of sensor-specific code enabling an ecosystem of reusable sensor drivers. The paper explores three alternative architectures for application-level drivers to understand trade-offs in performance, device portability, simplicity, and deployment ease. We explore these tradeoffs in the context of four sensing applications designed to support our work in the developing world. They highlight a range of sensor usage models for our application-level driver framework that vary data types, configuration methods, communication channels, and sampling rates to demonstrate the framework's effectiveness.

We present our experiences with an SMS-based system for providing transit information based solely on existing cellular and GPS networks. The aim is to permit the development of information services that do not rely on a central authority or complex web hosting. We developed and applied our system to the network of privately-run marshrutka buses in Bishkek, Kyrgyzstan. However, our goal is to more broadly address issues of ad-hoc shared transportation systems in the developing world. A custom designed GPS-GSM unit is placed on a vehicle, and users can query our server over SMS with their own non-GPS-enabled cell phones. We report on the accuracy of our location naming approach and estimates of bus arrival times. In addition, we summarize interviews with bus drivers and bus riders relating their views of the system and outline directions for future work. Our system is a grassroots solution to the persistent lack of transport information in developing countries.

This paper presents Open Data Kit (ODK), an extensible, open-source suite of tools designed to build information services for developing regions. ODK currently provides four tools to this end: Collect, Aggregate, Voice, and Build. Collect is a mobile platform that renders application logic and supports the manipulation of data. Aggregate provides a "click-to-deploy" server that supports data storage and transfer in the "cloud" or on local servers. Voice renders application logic using phone prompts that users respond to with keypad presses. Finally, Build is a application designer that generates the logic used by the tools. Designed to be used together or independently, ODK core tools build on existing open standards and are supported by an open-source community that has contributed additional tools. We describe four deployments that demonstrate how the decisions made in the system architecture of ODK enable services that can both push and pull information in developing regions.

This work consists of two main components: (a) a longitudinal ethnographic study in Kyrgyzstan that demonstrates the importance of transportation resources in the developing world and how to plan for an appropriate ICT solution, and (b) the results of a proof-of-concept system engineered to create a bottom-up, transportation information infrastructure using only GPS and SMS. Transportation is a very important shared resource; enabling efficient and effective use of such resources aids overall development goals.

Ubiquitous computing applications often use a user's context to automatically adjust their behavior to the situation. We have developed three types of wireless sensor nodes that can be worn, carried, or embedded in the environment that can provide interesting contextual information: a rich multi-sensor node to infer human activity that can be worn by a person or be part of a sensing environment, a portable wireless node for reading and writing short-range radio-frequency identification (RFID) tags, and a small wrist-watch-sized wireless display device that can serve as an easily accessible and 'glanceable' user interface. We report on our experiences in building these platforms and using them in some initial applications. We conclude with suggestions for future work in creating platforms that help blend sensor networks into ubiquitous computing.