Graduate School and Faculty of Electronic Information Systems, Systems Engineering and Science, Shibaura Institute of Technology.
My research focuses on developing software better, mainly for embedded systems such as Internet of Things (IoT), robots, and automobiles. My main tool is "models" such as UML, but I am also interested in using various methods to describe software. Recently, I have also been focusing on how machine learning can assist in day-to-day development.
Model-Driven Development (MDD) is a development method that mainly uses models such as UML to describe various aspects of software. Even if it is a model, I prefer a machine-readable and executable description rather than a human-readable description, and I think it is useful to have a description method that can be executed and verified. For example, if we draw a system using executable UML (xtUML), we can estimate the power consumption of the software created from it, and we are studying how to develop software that automatically runs adaptively within the available power budget.
We also focus on research using Domain-Specific Languages (DSLs), which use a language that is specific to a particular domain, since general languages are not sufficient to solve development problems. Development using DSL can reduce the amount of description and increase the quality. Our research focuses on security-aware DSLs for smart buildings and homes, and the environment to execute them securely.
We are also working on the automatic extraction of various models based on requirement documents and design documents written in natural language. These technologies are very effective in large-scale development.
Context-Oriented Programming (COP) uses the power of a new programming language to help develop software that behaves according to the context in which the system is placed. COP is still a new technology, and there is not enough knowledge in the community about how to use it well, especially for embedded systems. Therefore, we are researching how to apply COP to robot operating systems (ROS) that support robot development and how to guarantee real-time performance.
We are doing research on mobile applications and the Web that run on smartphones. We are working on an energy consumption profiler that visualizes the amount of power consumption per method and per screen component, power consumption analysis of various libraries that make up the mobile web, and aesthetic evaluation of mobile apps.