Abstract
In this paper, we analyse the effect of using appropriate tie breaking criteria in dispatch rules. We consider four different dispatch procedures, and for each of these heuristics we compare two versions that differ only in the way ties are broken. The first version breaks ties randomly, while the second uses a criterion that incorporates problem-specific knowledge. The computational results show that using adequate tie breaking criteria improves the performance of the dispatch heuristics. The magnitude of the improvement is different for the four heuristics, and also depends on the characteristics of each specific instance. The use of problem-related knowledge for breaking ties should therefore be given some consideration in the implementation of dispatch rules.

Abstract
This paper describes the study on example selection in regression problems using mu-SVM (Support Vector Machine) linear as prediction algorithm. The motivation case is a study done on real data for a problem of bus trip time prediction. In this study we use three different training sets: all the examples, examples from past days similar to the day where prediction is needed, and examples selected by a CART regression tree. Then, we verify if the CART based example selection approach is appropriate on different regression data sets. The experimental results obtained are promising.

Abstract
This paper discusses a new approach to the development life-cycle of agent-based production control applications, from the design to the operation, based in a catalogue of High-level Petri nets. The High-level Petri net-based approach facilitates the conception, definition and formal specification of an "encapsulation process" in industrial production systems. The catalogue includes elements for the identification of manufacturing components, the development of agent-based control units, the formal validation of the models and the formal specification of complete collaborative automation scenarios.

Abstract
This paper presents a hybrid tabu search strategy for lot sizing problems. This strategy allows the exploitation of the quality of the well-known relax-and-fix heuristic, inside a tabu search framework which enforces diversity. The computational results show an advantage of this strategy when compared to a version of the relax-and-fix heuristic and to time constrained branch-and-bound.

Abstract
In the last years, several manufacturing control architectures using emergent paradigms and technologies, such as multi-agent and holonic manufacturing systems, have been proposed to address the challenge of developing control systems capable of handling certain types of disturbances at the factory level. One of these holonic architectures is ADACOR, which integrates a set of paradigms and technologies for distributed manufacturing systems complemented by formal modelling techniques, to achieve a flexible and adaptive holonic/collaborative control architecture. The results obtained in the first experiments using the ADACOR architecture are presented in this paper, and also compared to the results produced by other control architectures. For this purpose a set of quantitative and qualitative parameters were measured, to evaluate static and dynamic performance of the control architectures.

Abstract
This paper presents a digital controller suitable for an instructional laboratory in electric drives. A prototype has been designed specifically to meet the requirements of low cost and motivating education tool, and it contains all of the active functions required to implement the open loop control of induction motors. This approach develops the concept of "grey box" module where the details are built for convenience, but not hidden inside as a "black box". Introductory laboratory experiments are presented to demonstrate how this hardware can be integrated in laboratory classes to present the most important concepts as regards adjustable speed drives.