Laboratory for System Dependability

We are actively accepting students pursuing master or doctoral degrees of computer science, international research students, and visitors for collaborative research on system dependability.
If you have interests, please contact us!

Our society is increasingly relying on ICT systems supporting smart transportation systems, power grid, financial, medical systems and so forth.
Since those ICT systems are indispensable for our lives, continuous engineering efforts to improve system dependability are imperatively important.
Our research laboratory tackles this issue by leveraging stochastic models and analysis techniques.
We model various uncertainties causing system misbehaviors such as component failures, estimate the system dependability quantitatively,
and evaluate the effectiveness of the measures to improve the dependability with lower cost.

Our studies are rooted on the belief in the necessity of DMAO cycle for continuous improvements of system dependability.

The cycle is composed of the set of processes for Design/Data collection, Modeling, Analysis and Optimization/Operation.
In this cycle, engineers design the information system and collect necessary data, model the designed systems with stochastic models,
analyze the model to quantify the dependability measures, explore the optimum solution to improve the dependability under the given constraints.
The results of optimization and/or real system operations should feedback to the system design repeatedly.

2020/11/16	Our paper entitled "Analysis of optimal file placement for energy-efficient file-sharing cloud storage system" has been accepted for publication in an upcoming issue of the IEEE Transactions on Sustainable Computing. It was a result of joint research work with Dr. Kazuhiko Kato, Dr. Koji Hasebe and Dr. Hirotake Abe in University of Tsukuba. [paper]
2020/10/13	As a Fast Abstract Co-chairs of ISSRE2020, we organized three Fast Panel and Abstract sessions: - Session 1: The emergence of machine learning in software reliability engineering - Session 2: The future of automated software reliability engineering - Session 3: New innovative approaches for system assurance
2020/10/12	Prof. Ermeson Andrade presented our co-authored paper in WoSAR2020. [paper] The data set obtained in our experiments for software aging in image classification systems are available. [data]
2020/8/22	Two papers co-authored by Dr. Fumio Machida have been accepted by WoSAR2020. - F. Machida and P. Maciel, "Markov chains and Petri nets for software rejuvenation systems" - E. Andrade, F. Machida, R. Pietrantuono, D. Cotroneo, "Software aging in image classification systems on cloud and edge"
2020/7/20	The deadline of full research paper submission to PRDC2020 is now extended to August 1st. Please consider to submit your contributions. The conference is postoponed to the next year, but all the accepted papers will be published in the proceedings of this year.
2020/7/12	We held sessions for introducing our research laboratory at the open campus for degree program in computer science. We are accepting a few students who are considering to enroll in the degree program next year and join our group. Please don't hesitate to contact us!
2020/5/1	"Handbook of Software Aging and Rejuvenation" has been published by World Scientific Publishing. Dr. Fumio Machida and Prof. Paulo Maciel contributed to Chapter 5 Markov Chains and Petri Nets.
2020/4/16	New master students, Sixiang Wang and Koushi Ryuu, have joined our group.
2020/2/26	As a part of organizing committee for IEEE/IFIP International Conference on Dependable Systems and Networks (DSN2020) , we are currently open the calls for Fast Abstracts. Please consider to submit your contributions to the DSN Fast Abstracts and Posters track.
2020/1/15	A joint research work with Prof. Xiaolin Chang, "Analyzing software rejuvenation techniques in a virtualized system: service provider and user views" has been published in IEEE Access. [paper]
2019/12/6	Dr. Fumio Machida gave a presentation at the 17th dependable system workshop in Japan (DSW2019). His presentation title is "Reliability analysis of machine learning systems using N-version models". We would thank to the valueable comments from the audience.
2019/12/3	Dr. Fumio Machida presented his paper at the IEEE Pacific Rim International Symposium on Dependable Computing (PRDC 2019)[paper][slides] Thanks to the participants of one of the last sessions of the conference!
2019/10/19	Dr. Fumio Machida made a presentation about the reliability of machine learning system using diverse classifiers at a workshop of the Operations Research Society of Japan.
2019/9/16	Our paper entitled "On the diversity of machine learning models for system reliability" has been accepted for presentation at the IEEE Pacific Rim International Symposium on Dependable Computing (PRDC 2019).
2019/8/28	Fumio Machida's co-authored paper "Analysis of Software Aging Impacts on Plant Anomaly Detection with Edge Computing" with Ermeson C. Andrade in UFRPE has been accepted by International Workshop on Software Aging and Rejuvenation.
2019/6/24	Dr. Fumio Machida presented a workshop paper at Dependable and Secure Machine Learning 2019 (DSML) co-located with International Conference on Dependable Systems and Networks (DSN).[paper][slides]
2019/4/10	Our paper entitled "N-version machine learning models for safety critical systems" has been accepted for presentation at the international workshop on Dependable and Secure Machine Learning.
2019/4/8	Dr. Fumio Machida made a presentation at the Workshop on Education and Practice of Performance Engineering (WEPPE) co-located with International Conference on Performance Engineering (ICPE) 2019. [paper][slides]
2019/2/1	Dr. Fumio Machida launched the laboratory for system dependabiliy in computer science department at University of Tsukuba. The laboratory is open to enthusiastic students and research collaborators who have interests on system dependability.

Department of Computer Science, University of Tsukuba
Address: 1-1-1, Tennodai, Tsukuba, Ibaraki 305-8573
E-mail: machida at cs.tsukuba.ac.jp