Continuum Computing Trustworthiness Research Group

✪ Trust management of continuum networks

Our first focus is to improve the trustworthiness of continuum networks. We develop methods for the trust management of networks, especially on network security management and operation.

✪ Quality management of machine learning systems

Our second focus is on the quality of machine learning systems. We develop methods to improve and evaluate the implementations of machine learning algorithms, models, and systems from a software engineering perspective. We also develop the "Machine Learning Quality Management Guideline" to establish quality goals and development processes for products and services using machine learning.

• Machine Learning Quality Management Guideline (AIST Committee for Machine Learning Quality Management)

  • 3rd English Edition (January 2023)
  • 4th Japanese Edition (December 2023)

• International standardization

  • Activity on ISO/IEC TR 5469:2024

• Machine Learning Quality Management Project (NEDO funded project)

  • Open testbed toolset Qunomon for the quality management of AI systems

✪ Formal methods for software systems with uncertainty

Our third focus is to evaluate and certify the trustworthiness of software systems with uncertainty, such as cyber-physical systems. We develop formal methods for modeling and verifying software systems that deal with probabilistic events, physical environments, and so on. We also conduct foundational research on programming languages and interactive theorem provers.

• Formal verification of software and mathematics: program verification, program generation, mathematics, information security, robotics (in collaboration with Inria, Nagoya University, and others)

• Integration of formal methods and statistical methods: JST PRESTO, French-Japanese project LOGIS

• Education in Nara Institute of Science and Technology (NAIST) Formal Verification Lab