The integration of techniques from AI and OR has shown in the past years how it improves algorithm quality for tackling complex and large scale combinatorial problems, especially in terms of efficiency, scaleability and optimality. The value of this integration has been shown in applications such as hoist scheduling, rostering, dynamic scheduling and vehicle routing. At the programming/modelling level, most constraint languages embed OR techniques to reason about collections of constraints, so-called global constraints. A few also provide support for hybridization allowing the programmer to build new integrated algorithms. The resulting multi-paradigm programming framework combines the flexibility and modelling facilities of constraint programming with the special purpose and efficient methods from Operations Research.

The CP-AI-OR workshop is becoming a major forum for exchanging ideas and methodologies from both fields, as demonstrated by its success in Ferrara (1999), Paderborn (2000), Ashford (2001), and Le Croisic (2002). Its aim is to bring together interested researchers from AI and OR, and to give them the opportunity to show how the integration of techniques from AI and OR can lead to interesting results on large scale and complex problems. We explicitly welcome new ideas and methods for integrating OR and AI techniques that have arisen from real-world applications.

The 2003 workshop is organized by the Centre for Research on Transportation and will take place in Montréal, Canada on May 8-10 immediately following Optimization Days, which is a major OR conference held each year in Montréal. Arrangements will be made to facilitate attendance to both events and thus encourage discussions with the North American OR community.