The broader impact/commercial potential of this Small Business Innovation Research (SBIR) Phase I project involves Web, Mobile and IoT applications, which generate significant economic value for society. IoT applications alone will add about $2T in economic value over the next 10 years. However, in order to fully realize the economic potential, high value applications will need to be created and brought to the market at low cost and quickly. This project will develop products that will significantly reduce development cost, improve quality, and bring these applications to market faster. The proposed project will also have significant impact on computer science education. Given the growing interest in programming at high school and college levels, the project will produce software tools and educational material that students can use to build novel applications without needing to learn a wide variety of languages and tools to experiment with different mobile and IoT devices. Further, interest in IoT has given rise to the maker movement, where low cost software and hardware tools are being used to build smart objects. Tools produced by this project will further lower the barrier of entry for the maker community. This Small Business Innovation Research Phase I project addresses the problem of developing IoT, Mobile and Web applications. Developing these applications is extremely challenging because of the differences in device capabilities, platform support (iOS vs. Android vs. Windows), and resources (memory, CPU, power). Building such applications often has meant rewriting a large percentage of the application code using different programming languages, tools, libraries and platform support. This has resulted in high development and management costs, inconsistent and poor product experience across platforms, and low software quality. This project addresses the problem by using a language-driven approach. In this approach, a new language is designed specifically to provide abstractions for the various forms of heterogeneity (including hardware, capability, resource and abstraction heterogeneities). Compilers and code generators use the semantic properties of the abstractions to generate code customized for each device. The Phase 1 project will develop tools and libraries for targeting the language for several mobile and IoT platforms. The outcome of Phase 1 will be a set of tools, which when given a program and a description of devices, will generate code for each device, distribute and install the binaries on each device, and manage the binaries for the devices.
