Saddek Bensalem

Engineering Trustworthy Learning-enabled Autonomous Systems
(under construction)!

Saddek Bensalem

Professor, Universirsité Grenoble Alpes (UGA)

Contact

Phone: +33 (0)4 38 78 28 06

Fax: +33 (0)4 56 52 03 44

Saddek.Bensalem[at]univ-grenoble-alpes.fr

Laboratoire Verimag

Office location:
Bâtiment IMAG, Université Grenoble Alpes
700 avenue Centrale Domaine Universitaire - 38401 St Martin d'Hères
France

Research interests

I have been leading the "Rigorous System Design" research team, from 2008 to 2020, around the BIP project, with two main directions: 1) the development of verification methods and tools that scale and are usable by engineers ; 2) the design of embedded systems, with an emphasis on the formalization of the design process and the techniques that guarantee correctness by construction. The ambition of the work is to overcome theoretical limitations inherent in the complexity of monolithic verification methods and to conduct a holistic study of the problem of correctness in the general context of a rigorous and mathematically sound design. The BIP framework has found several application domains in practice, with excellent results (robotics, embedded, cyber-physical systems, etc.).

My research activity focuses on rigorous system design as a coherent and accountable process aimed at building cost-effectively systems of guaranteed quality. The aim is to provide the theoretical underpinnings, methods and tools for moving from empirical approaches to a well-founded discipline. Since October 2020, I created a new research project on "Engineering Trustworthy Learning-enabled Systems". The aim of the project is to develop foundations for continuous engineering of trustworthy LE-systems. The targeted scientific breakthrough lies within the convergence of "data-driven" and "model-based" engineering, where this convergence is further complicated by the need to apply verification and validation incrementally and avoid complete re-verification and re-validation efforts. The three scientific directions of this project are: (1) integration of learning-enabled components and model-based components via a contract-based methodology which allows incremental modification of systems including threat models for cyber-security, (2) adaptation of verification techniques applied during model-driven design to learning components in order to enable unbiased decision making, and finally, (3) incremental synthesis techniques unifying both the enforcement of safety properties as well as the optimization of performance.

Thematic mobility

Geographical mobility

I was a researcher on leave at CNRS from 1998 to 2000 and at CEA from 2012-2014. I spent several months as invited researcher at universities and research institute such as : SRI (sabbatical year in 1996, 3 months in 1997, 8 months in 1998 and 6 months in 1999), Univ. of Kiel (1 month in 1997, 98 and 99), Stanford University (2 months in 1998, 3 months in 2001and 2 months in 2006), Institute Weizmann (2 months in 1999), Univ. of Uppsala (2 months in 2001), Nasa Ames (3 months in 2003), Urbana Champaign Univ. (3 months in 2005), Nasa/JPL (1 month in 2013, 2014 and 2015), Bar Ilan Univ. (1 month in 2013), NIST (4 months in 2016 and 3 months in 2017).

Current Projects

Past Projects

Recent publications [all] [dblp] [google scholar]