Humbert Fiorino

Printer-friendly versionPDF version

Contact Information

Office 364
Laboratoire d'Informatique de Grenoble (UMR 5217)
Bâtiment IMAG, 700 avenue Centrale
Domaine Universitaire - 38401 St Martin d'Hères
Phone: +33 4 57 42 14 95
Email: Humbert.Fiorino@imag.fr

 

Research Interests

I am interested in devising autonomous agents (softbots, physical robots etc.). Agent technologies envision artificial systems with a high degree of autonomy requiring only high-level supervision of human users. This autonomy simplifies human control and drastically improves the scope of agent achievable tasks. Fully autonomous agents are still out of reach: currently their performances under human remote control are very limited. However, in the near future, autonomous agents will become believable companions in our everyday lives and useful surrogates in dangerous or troublesome situations. The concepts and the algorithms that I investigate in my research works aim at contributing to the development of "autonomy technologies". This research effort is important to support the incipient industry of autonomous systems.

Generally, autonomy in artificial systems is considered as a restrained form of rational reasoning for action, i.e. the ability for an agent to make its own sequential decisions in order to achieve an assigned objective i.e. automated planning. I have addressed the autonomy of artificial systems in other dimensions:

  • Composition: This is the ability for an agent to fulfill objectives that are beyond the scope of its intrinsic capabilities: it involves a cooperative encounter based on distributed planning between agents which jointly explore a shared search space, identify and look for necessary resources, coordinate their decisions and operate the appropriate mechanisms to solve their conflicts.
  • Generalization: This is the ability to compute high-level/algorithm-like plans that solve classes of planning problems. Generalized planning allows to expand agents' capabilities by identifying non deterministic choices, concurrence and loop-structures in planning problems.
  • Adaptation: This is the ability to learn from past decisions, and explore search spaces with respect to everchanging goals in dynamic environments.

Short Biography

I received my Ph.D. degree from SUPAERO (Ecole nationale supérieure de l'aéronautique et de l'espace) and the ONERA (Office National d'Etudes et Recherches Aérospatiales). I also visited the Robotics and Artificial Intelligence group at the LAAS (Laboratoire d'Analyse et d'Architecture des Systèmes) as postdoctoral fellow. This research work was supported by the DGA (Délégation Générale à l'Armement) and focused on AUV (Autonomous Underwater Vehicles) cooperation. Currently, I am Associate Professor in computer science at Université Joseph Fourier and permanent researcher at LIG.