
Logic and Artificial Intelligence
research
group at
Main research
line:
Methods and foundations of logic-based
tools for dealing with hard
computational problems
Logic
plays a fundamental role in many Computer Science areas:
programming languages, databases, software engineering,
software and hardware verification, etc.
The use of mathematical logic as a formalism for Artificial
Intelligence was recognized by John McCarthy as early as in
1959, and since then it has played an important role in some central areas of AI research. Logic has
also been called "the calculus of computer
science", as
it plays a similar role in Computer Science to that of
calculus in the
physical sciences and traditional engineering disciplines (M.
Vardi,
2007)
The group's research is related with automated
reasoning (automated theorem proving in
first-order logic with equality, higher-order
unification, and multiple-valued logics) as well as to
propositional satisfiability
(SAT) and satisfiability modulo theories (SMT) and their applications to
solve combinatorial
problems in AI areas such as planning and scheduling.
Faculty
-
-
Francesc Castro
-
Josep Suy
-
PhD
Students
-
Jordi Coll, PhD (grad. 2019)
-
Joan Espasa, PhD (grad. 2018, later at Launchmetrics)
-
Miquel PalahÃ, PhD (grad. 2015, later at Unipart Group)
Graduate
Students
-
Marc Garcia
-
Gerard MartÃn
Software
-
WSimply:
a system for solving constraint satisfaction problems
using SMT.
-
fzn2smt:
a compiler from the FlatZinc language to the standard
SMT-LIB language.
-
rcpsp2smt:
a system for solving the Resource-Constrained Project
Scheduling Problem using SMT.
-
mrcpsp2smt:
a system for solving the Multi-mode Resource-Constrained
Project Scheduling Problem using SMT.
-
rcpspt2smt:
a system for solving the RCPSP with Time-Dependent
Capacities and Requests using SMT.
-
mrcpspmax2smt:
a system for solving the Multi-mode Resource-Constrained
Project Scheduling Problem using SMT.
-
viewproject:
a system for editing Resource-Constrained Projects
-
TILC:
the interactive lambda-calculus tracer.
-
RanTanPlan:
A Planner based in the planning as satisfiability
approach, that relies on a SMT solver.
-
B2B benchmarks:
dzn files used in the experiments of the paper
"Constraint Solving Approaches to the Business-to-Business Meeting
Scheduling Problem" (to appear)
Publications: see
personal pages
Last update: September 30, 2021