Analysis, Design and Management of manufacturing systems
Research focus
The Peer review has evaluated this group as Excellent
The activities in this research topic take the move from on the analysis of the evolution of the European and worldwide context in which manufacturing systems operate and aim at the definition of new solutions to production problems. Research in this direction is underpinned by three main pillars in which specific research and innovation is carried out namely: Performance evaluation, Production system configuration, Production planning. 1. Performance evaluation The activities in this domain aim at the assessment of the performance of production systems considering the dynamic interaction among the devices composing the system. The solution proposed are based on virtual manufacturing at machine level [ADM-RP9][ADM-BC6], while at system level approximate analytical methods [ADM-RP1/3/4/6/8][ADM-P2] [ADM-BC4], simulation [ADMRP5][ ADM-P7/8], and perturbation analysis are developed and applied. Approximate analytical methods are new techniques based on Markov Chains and Queuing theories allowing the estimate of the performance of the systems without the restrictive hypotheses of classical analytical techniques. Perturbation analysis is a numerical technique to estimate the sensitivity of the main system performance in terms of gradients of production rate and lead times. Discrete event simulation models and techniques are developed for very detailed estimation of system performance and are also used as a comparison mean for the new techniques developed. An integrated use of these techniques provides an accurate and complete set of estimated performance indicators to be used during the analysis, design and management of production systems. 2. Production system configuration Products evolve during their lifecycle, by changing their technological characteristics and therefore their requirements on the production system. Moreover the introduction of new products and product variants, the structural modification of production volumes requested by the market introduce additional variability on the requirements the production system must satisfy. As a consequence, production systems may be forced to run inefficiently because they are no longer consistent with the modified production requirements. To overcome this problem, possible future scenarios must be considered from the very beginning and the configuration activity must introduce the appropriate levels of flexibility and reconfigurability in the proposed system configuration. The activities in this domain provide solutions to select the most appropriate configuration of production system that closely matches the product lifecycles of 94 actual and future products. Thus solutions are developed to make appropriate selections among production system architectures (transfer lines, flexible manufacturing systems, adaptable manufacturing systems, machining cells or centres, modular systems) [ADM-RP2][ADM-P3/9][ADM-BE1], production resources (machines, tools, transporters, fixtures, etc.) [ADM-RP7] and resource allocation (number of elements for each production resource) [ADM-P6/7][ADMBC2/ 3/5] in order to address the production problems of a firm in present and future contexts. These solutions are based on optimization techniques such as mixed integer linear programming, stochastic programming, dynamic programming, response surface methodology and meta-heuristics such as tabu search, simulated annealing and genetic algorithms. 3. Production planning Despite the manufacturing environments are in great part characterized by uncertainty, most production planning approaches assume perfect information and a static deterministic environment. The occurrence of uncertain events can however have a significant impact on the stability and the performance of the production system by affecting the meeting of due-dates, the efficient resource allocation and the usage of non-regular working force. Unexpected events can stem from sources internal or external to the manufacturing system: duration of activities, availability of production resources, delivery of raw materials, occurrence of new activities, modifications of release dates or due dates. The main goal in this domain is to devise new solution to the production planning problem able to guarantee robustness against the high level of uncertainty characterizing the manufacturing environments [ADM-RP5]. Robust production planning must provide a schedule of the activities and of the utilisation of resources incorporating a certain degree of anticipation of uncertainty. Robustness of the production plan relies on the possibility of react to unexpected events with as little as possible impact on the performance of the production system and on the inbound and outbound supply chain [ADM-P12/13]. The research activities comprehend the development of new approach for robust production planning, stochastic modelling of uncertainty in manufacturing environments and characterisation of production plans in terms of robustness properties. The adopted methodologies are mixed integer linear programming, stochastic programming and constraint programming together with dedicated scheduling optimisation algorithm and heuristics techniques. All the described methods and models are developed with specific refence to industrial cases and many of them have been applied in industry with particular reference to the machine tool sector and the mechanical sector. Manufacturing system considered include assembly/disassembly systems, FMSs, job shops and new manufacturing systems architectures.
Dipartimento di afferenza
Docenti afferenti
Full Professors
Francesco Jovane
Quirico Semeraro
Tullio Tolio
Assistant Professors
Marcello Colledani
Andrea Matta