Breakthroughs in molecular and synthetic biology through de novo gene synthesis are stimulating new vaccines, pharmaceutical applications, and functionalized biomaterials, and advancing the knowledge of the function of cells. This evolution in biological processing motivates the study of a class of manufacturing systems, defined here as genetic manufacturing systems, which produce a final product with a genetic construct. Genetic manufacturing systems rely on rare molecular events for success, resulting in waste and repeated work during the deoxyribonucleic acid (DNA) fabrication process. Inspection and real time monitoring strategies are possible as mitigation tools, but it is unclear if these techniques are cost efficient and value added for the successful creation of custom genetic constructs. This work investigates resource allocation strategies for DNA fabrication environments, with an emphasis on inspection allocation. The primary similarities and differences between traditional manufacturing systems and genetic manufacturing systems are described. A serial, multi-stage inspection allocation mathematical model is formulated for a genetic manufacturing system utilizing gene synthesis. Additionally, discrete event simulation is used to evaluate inspection strategies for a fragment synthesis process and multiple fragment assembly operation. Results from the mathematical model and discrete event simulation provide two approaches to determine the appropriate inspection strategies with respect to total cost or total flow time of the genetic manufacturing system. / Ph. D. / Breakthroughs in molecular and synthetic biology through <i>de novo</i> gene synthesis are stimulating new vaccines, pharmaceutical applications, and functionalized biomaterials, and advancing the knowledge of the function of cells. This evolution in biological processing motivates the study of a class of manufacturing systems, defined here as genetic manufacturing systems, which produce a final product with a genetic construct. Genetic manufacturing systems rely on rare molecular events for success, resulting in waste and repeated work during the deoxyribonucleic acid (DNA) fabrication process. Inspection and real time monitoring strategies are possible as mitigation tools, but it is unclear if these techniques are cost efficient and value added for the successful creation of custom genetic constructs. This work investigates resource allocation strategies for DNA fabrication environments, with an emphasis on inspection allocation. The primary similarities and differences between traditional manufacturing systems and genetic manufacturing systems are described. A serial, multi-stage inspection allocation mathematical model is formulated for a genetic manufacturing system utilizing gene synthesis. Additionally, discrete event simulation is used to evaluate inspection strategies for a fragment synthesis process and multiple fragment assembly operation. Results from the mathematical model and discrete event simulation provide two approaches to determine the appropriate inspection strategies with respect to total cost or total flow time of the genetic manufacturing system.

La compétitivité d'une entreprise est conditionnée par l'aptitude de son organisation de trouver des solutions pour améliorer l'efficacité de son processus de fabrication en maîtrisant sa variabilité tout en garantissant des coûts bas, des délais réduits et, parfois, une certaine capacité de produire à des grands volumes. La maîtrise du processus de fabrication est un ensemble d'activités réalisées par un processus opérationnel de contrôle en suivant un plan de surveillance préétabli avec des objectifs précis en termes de maîtrise des risques. La mise en application du plan de surveillance est souvent mise à mal par : les aléas liées aux flux physiques et informationnels, l'interaction du processus de contrôle avec d'autres processus opérationnels de l'entreprise et les limitations en ressources de maîtrise. L'objectif de cette thèse était de répondre à cette problématique en proposant des approches nouvelles pour la conception des plans de surveillance. Pour cela, nous proposons une approche assurantielle permettant de définir un plan de surveillance qui limite l'exposition aux risques et optimise l'utilisation des ressources de maîtrise. Une formalisation simplifiée a permit d'expliquer cette approche à travers des exemples. Les résultats des expérimentations sur un cas industriel, montrent l'intérêt de l'approche pour la maîtrise des risques opérationnels dans un processus de fabrication. Par la suite, une formalisation généralisée de l'approche a été proposé concevoir le plan de surveillance généralisé d'un processus disposant d'un ensemble des moyens de contrôle interposé sur plusieurs couches le long du processus qui est assujetti à multiples risques. Mots clés : Plan de surveillance, Maîtrise des risques, Allocation des inspections, Capabilité, Capacité, Performances. / The competitiveness of a company depends on its organization ability to improve the efficiency of the manufacturing process by controlling its variability while maintaining low costs, reduced delays and, sometimes, a certain capacity to produce at high volumes. Process monitoring is a set of activities performed by an operational process of control according to the control plan which is predetermined with specific goals in terms of risk control. The implementation of the control plan is often undermined by: the uncertainties associated with physical and informational flows, the interaction of the control process with other operational processes and the limitations of control resources. The objective of this thesis was to address this problem by proposing new approaches to the design of control plans. For this, we propose an approach for defining an insurance-based control plan that limits exposure to risk and optimizes control resources utilization. A simplified formalization has allowed explaining this approach through examples. The results of experiments, done on an industrial case study, demonstrate the value of the approach for operational risk management in a manufacturing process. Subsequently, a generalized formalization of the approach has been proposed to design the control plan of a generalized process with a set of control resources interposed in several layers along the process which is subject to multiple risks. Keywords : Control plan, Risks monitoring, Inspection allocation, Capability, Capacity, Performances.