MINES Saint-Étienne - Inspiring innovation
LinkedIn Youtube Instagram Research Gate Google Scholar
UK
Rechercher
Close
Menu

Research

Home | Research | Centres and Departments | Manufacturing Sciences and Logistics Department (SFL) | PhD Theses

PhD Theses

Current theses

Optimized disruption recovery in liner shipping

  • Ph.D. Student: M. MERLE
  • Grant: CIFRE CMA CGM
  • Start: March, 2024
  • Advisors: N. Absi, D. Feillet
  • Description

    When an operational disruption occurs, shipping companies have to tackle a complex decision-making problem when it comes to recover the schedule of the affected vessel. Once the rescheduling solution has been approved, a new problem arises : rerouting the containers. An evacuation plan must then be found for the containers that cannot be transported under the conditions initially planned.

    The work in this thesis will focus on the modelling of decision problems linked to operational disruptions encountered in the maritime industry. The final objective of the thesis is to provide CMA CGM employees with robust decision support tools based on combinatorial optimization algorithms. These tools will enable employees to make decisions more efficiently, while being aware of the implications in terms of cost, time, and environmental impact.

Planification horaire adaptative des trains SNCF

Planification intégrée et robuste des ressources ferroviaires pour le transport de marchandises

  • Ph.D. Student: L. FOURCADE
  • Grant: Industrial research agreement (CIFRE) with STMicroelectronics
  • Start: May, 2022
  • Advisors: S. Dauzère-Pérès
  • Description

Semantic Modeling for the industry 4.0

  • Ph.D. Student: G. MASLOV
  • Grant: Industrial research agreement (CIFRE) with STMicroelectronics
  • Start: Sept, 2021
  • Advisors: C. Yugma
  • Description

    Defining a flexible methodology for the development of conceptual models (semantic modeling) capable of ultimately covering the entire manufacturing scope for future industry 4.0 applications such as production control, data analysis or knowledge management.

Optimal architecture of R2R control loops and multivariate predictive models including operational constraints as a support to the digitalization of the microelectronics industry

  • Ph.D. Student: L. TERRAS
  • Grant: Industrial research agreement (CIFRE) with STMicroelectronics
  • Start: Aug, 2021
  • Advisors: A. Roussy
  • Description

    Driven by the demand of process miniaturization, the semiconductor industry is getting more and more complex. Nowadays, advanced technology products are processed through more than a thousand dependent and highly sophisticated operation steps. In addition, the permanent re-entrant flows make the products to be processed on any available equipment: considering one operation, products sharing the same recipe can be processed on different tools, and different products can be processed on the same tool. The development of a Run-to-Run (R2R) control system able to efficiently compensate process variations is at the stake of the industry 4.h priorities. The objective of this thesis is to develop a new R2R algorithm based on multivariate predictive models in order to improve the efficiency of the existing system. The performance of the proposed approach will be analyzed in term of variability reduction, sampling robustness and easy maintenance.

Robust scheduling within SNCF railway maintenance centers

  • Ph.D. Student: R. TORBA
  • Grant: Industrial research agreement (CIFRE) with SNCF
  • Start: May, 2021
  • Advisors: C. Yugma, S. Dauzère-Pérès
  • Description

    The SNCF industrial technicentres carry out heavy maintenance of its rolling stock. The aim is to renovate and modernize railway vehicles and their components every 10 to 20 years in order to increase their life cycle since the costs of rolling stock are very high. Heavy maintenance operations are very complex to be precisely estimated in advance and hence there are many sources of uncertainty (additional work, accidental repairs, uncertainties of operating times estimation, machine breakdowns etc.). The objective of this thesis is to explore robust scheduling algorithms to guarantee schedules with good performance even in the event of disturbances. The probability of respecting weighted client release dates but also the minimization of weighted rolling stock crossing times will be used to evaluate the performance of these algorithms. Historical and empirical data will be used to build different scenarios, to measure the robustness and validate the proposed approaches.

AI for the optimization of port logistics

  • Ph.D. Student: D.SERHAL ABOU NADER
  • Grant: Industrial research agreement (CIFRE) with DMS LOGISTICS
  • Start: May, 2021
  • Advisors: D. Feillet, N. Absi, Thierry Garaix
  • Description

    The intermodal transportation has emerged greatly in the 20th century because of its speed and cheapness in transporting goods. While great researches have been made to optimize the maritime transportation of containers, the land-based areas are still far behind from being optimal which is affecting all the supply chain actors. This is practically translated by late deliveries, long waits at terminals and congested ports, and incurring huge increases in CO2 emissions as well as massive gain losses. Therefore, the main challenge is to achieve a global vision of logistics in order to anticipate the various container flows and to guide all the parties in a supply chain in the conduct of their operations. In this context, the aim of this PhD is to provide the different actors of container logistics with a tool for decision-making support and to enable visibility on the land operations of containers to optimize and fluidify their activity. More specifically, the objective is to understand the mechanisms at the origin of slowdowns in container exchanges, and to be able to formulate prediction and optimization algorithms.

Optimization and Simulation for an On Demand Road-Rail Transport in Few Populated Area

  • Ph.D. Student: J. JODEAU
  • Grant: Industrial research agreement (CIFRE) with SNCF
  • Start: May, 2021
  • Advisors: D. Feillet, N. Absi
  • Description

    In sparsely populated area public transport are insufficient and do not suit users’ demand. There is high-capacity train circulating with a low frequency following a predefined timetable. Therefore, users rather use their own car than public transport for its ease and flexibility. To revitalize public transport and rail transport in these areas a mixt on-demand transport based on an innovating rail-road vehicle is considered. This solution aims at transporting the most passenger in door-to-door transportation during a specified period while minimizing the overall user’s travel time. Vehicles circulate on a single-track rail line and make some detour to pick up or drop off users.

Tactical and Operational Management of Time Constraint Tunnels

  • Ph.D. Student: B. Anthouard
  • Grant: Industrial research agreement (CIFRE) with STMicroelectronics
  • Start: May, 2021
  • Advisors: S. Dauzère-Pérès
  • Description

    Semiconductor wafer fabrication includes one of the most complex and constrained manufacturing processes, involving a highly cyclical and large set of operations with long cycle times. For the sake of product yield and quality considerations, time constraints (TCs) are prescribed between process operations with varying degrees of rigor. With the shrinkage of pattern sizes, the increased sensitivity of wafers leads to a growing number of time constraints. The resulting multitude of time constraints overlap and follow each other in close succession, forming the so-called time constraint tunnels (TCTs). TCTs are nowadays one of the main hot points for both production planning and day-to-day line management. The proposed Ph.D. thesis aims to develop competitive approaches for the TCT management, by addressing tactical and operational issues:
    . Tactical level: To study and control the impact of TC duration on mid-term key performance indicators (e.g. production flow, throughput rate, Work-In-Process thresholds, capacity);
    . Operational level: To propose efficient control mechanisms (e.g. lot gating mechanism, Work-In-Process positioning, preventive maintenance and recipe validation management) to be integrated into real-time decision-support tools operating in highly time-varying environments.

Integration of process variability for robust scheduling in complex workshops

  • Ph.D. Student: M. Flores Gómez
  • Grant: IPCEI project Nano2022
  • Start: March 2020
  • End: June 2023
  • Advisors: S. Dauzère-Pérès, V. Borodin
  • Description

    This thesis aims to develop and validate on industrial data, modern approaches to incorporate uncertainty and process variability in scheduling problems specific to complex manufacturing workshops. A robust schedule is important to ensure the stability of the production system and account for process variability. This variability is associated to the uncertainty of setup and processing times on the machines, machine availability, uncertain job releases and other factors. We focus on the variability of processing times and their impact on schedule’s performance. The goal is to adapt existing algorithms using performance measurements in order to generate robust schedules, capable of facing processing times variability.

Traffic management of a mixed CBTC suburban railway line

  • Ph.D. Student: H. MEUNIER
  • Grant: Industrial research agreement (CIFRE) with SNCF
  • Start: March 2020
  • End: June 2023
  • Advisors: S. Dauzère-Pérès, V. Borodin
  • Description

    To improve the performance of complex railway lines in highly dense areas, automatic train control systems are increasingly being implemented in railway networks, such as the Communication-Based Train Control (CBTC) system. This system makes suburban line more efficient but also more sensitive. Hence, real-time and automated rescheduling decisions are needed. More precisely, the aim of this PhD thesis is to provide real-time and automated rescheduling decisions for a mixed and open suburban line, equipped with a CBTC system, in a very dense area. Using the position and the state of trains, a multi-objective optimization module proposes a new timetable. Headway and punctuality objectives are targeted while perturbations occur. The new timetable is applied by providing to each train new arrival and departure times, travelling speed and paths to use. Specific efforts are made to propose feasible short-term solutions by considering practical aspects such as headway and signaling rules. Tests are conducted using different disturbance scenarios on the simulated infrastructure of the future tracks of the E line of the Paris suburban rail network, using the simulation engine SIMONE.

Optimization and machine learning for multi-objective scheduling in complex manufacturing areas

  • Ph.D. Student: J. Berthier
  • Grant: Industrial research agreement (CIFRE) with STMicroelectronics
  • Start: March 2020
  • End: September 2023
  • Advisors: S. Dauzère-Pérès, C. Yugma
  • Description

    The semiconductor sector is one of the most complex and competitive industry with particularly long, intricate and costly manufacturing processes. Controlling production lead times is a major issue in order to balance between manufacturing costs, product obsolescence and compliance with customer deliveries. In this context, scheduling products on equipments at the various stages of their manufacturing route has become a  crucial decision to make.

    This thesis aims to extend scheduling approaches already used at STMicroelectronics in order to better integrate existing objectives, model new criteria and consider new manufacturing areas. The resulting scheduling engine will embed an automatic adjustment of the optimization parameters to the workshop context based on machine learning techniques. Academic contributions are expected in the fields of multi-objective optimization and statistical learning methods, which are both of increasing interest to the scientific community.
  • Keywords: industrial engineering; operations research; scheduling; multi-objective optimization; machine learning; semiconductor industry

Human-aware tactical planning problems in warehouse management

  • Ph.D. Student: T. Prunet
  • Grant: ANR projet AGIRE
  • Start: February 2020
  • Advisors: N. Absi, V. Borodin, D. Cattaruzza
  • Description

    This thesis focus on the intersection of two trending OR research streams: warehousing and ergonomics. Warehouses are one of the key components of any supply chain, with several problems that have been well studied in the literature. Despite the recent technological advances in automation, most warehouses still rely on human labor, for a work that is prone to the apparition of work-related muscolulo-skeletal disorders. Corporate Social Responsability and employee well-being are becoming more and more a concern for companies. This global trend has been studied in a relatively recent, and dynamic, stream of research in OR about the integration of Human Factors and Ergonomics (HF/E) in decision models. The topic of this thesis is to investigate, and develop new models, for the integration of HF/E into tactical planning problems occurring in warehouse management.

Development of a multi-source and multi-level data aggregation system to assist the diagnosis of process controls in real-time

  • Ph.D. Student: I.RABHI
  • Grant: CIFRE industrial research agreement with SNCF
  • Start: January 2020
  • End: June 2022
  • Advisor: Agnès Roussy
  • Description

    In semiconductor industry, Statistical Process Control is a mandatory methodology to maintain under control all the process steps and to guarantee a high level of quality. It has two main objectives: the detection of out-of-controls and the identification of potential root causes in order to correct them. The objective of this thesis is to improve the classical SPC system currently adopted, by first proposing a machine learning based technique that will allow to exploit the full metrology raw data in order to better detect anomalies on the wafers. Then identifying their potential root causes, using a multidimensional data analysis based on all the historical data of the process steps preceding the measurement that represented an anomaly. By doing this, we will have an SPC 4.0 adapted to today’s advanced technologies with complex processes.

Bi-level Multi-objective Optimization Problems using Evolutionary Algorithms

  • Ph.D. Student: M.ABBASSI
  • Grant: Academic
  • Start: December 2019
  • End: September 2022
  • Advisor: N. Absi, Lamjed Ben Said
  • Description

    Many real-life applications are modelled using a hierarchical decision-making in which: an upper-level optimization task is constrained by a lower-level one. Such class of optimization problems is referred in the literature as Bi-Level Optimization Problems (BLOPs). Most of the proposed methods tackled the single-objective continuous case adhering to some regularity assumptions. This is at odds with real-world problems that involve mainly discrete variables and expensive objective function evaluations. Besides, the optimization process becomes exorbitantly time consuming, especially when optimizing several objectives at each level. For this reason, the Multi-objective variant (MBLOP) remains relatively less explored and the number of methods tackling the combinatorial case is much reduced. Motivated by these observations, we propose in this PhD thesis Evolutionary Algorithms (EAs) to solve combinatorial MBLOPs. We propose generic EAs to facilitate the decision making for this kind of application and particularly we tackle the production distribution collection and disassembly processes in the supply chain.

Optimization and machine learning for the rescheduling and the design of transportation plans in a dense railway system

  • Ph.D. Student: H. Belhomme
  • Grant: CIFRE industrial research agreement with SNCF
  • Start: December 2019
  • End: July 2023
  • Advisor: S. Dauzère-Pérès, D. Feillet
  • Description

    In the dense railway system of the Paris area, small delays can easily propagate and disturb the stability of the network, potentially leading to significant perturbations. This situation constitutes a large, real time, multi-objective rescheduling problem. Two main objectives are opposed in this multi-objective problem: To minimize the duration of the disruption and to minimize the impact on the service to passengers. This thesis aims at improving a decision support tool for the managers of the operational management centers. The work is structured along 3 axes: (1) Improving the multi-objective aspects of the decision support tool, (2) Pairing a machine learning method with the optimization method and (3) Proposing an approach to improve the design of the transportation plans.

Integration of production scheduling and transportation for industry 4.0

  • Ph.D. Student: L.Berterottiere
  • Grant: Bourse Mines
  • Start: Decembre, 2019
  • Advisors: C. Yugma, S. Dauzère-Pérès
  • Description

      More and more production units are now able to forecast production, particularly with the increasing use of scheduling decisions in all aspects of production. We are moving from reactive transportation and storage management to predictive management where these operations can be planned. Most existing problems in the literature focus on routing or scheduling, they are rarely integrated and still more rarely they include storage management. The purpose of this PhD research is to integrate scheduling, routing, and storage management to make better decisions.

Development of a generic and industrially viable management system for metrology tool control and production regulation

  • Ph.D. Student: R. Clain
  • Grant: European project MadeIn4
  • Start: December 2019
  • End: May 2023
  • Advisor: A. Roussy, V. Borodin
  • Description

    Metrology tools are designed to perform identical measurements of a set of relevant manufacturing parameters to help controlling the production process. In practice, measures provided by the same metrology tool can be different for two identicD. Feilletal samples, which should not be. The metrology and inspection stages are particularly challenging in semiconductor manufacturing systems, due to the presence of reentrant flows, process tool heterogeneity, and High-Mix Low-Volume (HM-LV) production configurations. The ability to perform a consistent exploratory analysis and to provide a valuable support across all the tools is highly critical for semiconductor manufacturing. To respond to these industrial issues, this thesis aims to develop an industrially viable and generic methodologies to efficiently use metrology measurements for supporting the metrology tool control and production regulation. This thesis will be carried out in close cooperation with STMicroelectronics Crolles and STMicroelectronics Rousset within the European Project MadeIn4.

Equity in mobility systems for home services

  • Ph.D. Student: M. Agius
  • Grant: ANR project FITS
  • Start: June 2019
  • End: September 2023
  • Advisor: N. Absi, D. Feillet, T. Garaix (CIS)
  • Description

    Mobility systems consist in scheduling routes for vehicles to serve client requests. It is one of the most studied problems in Operations Research, and is known as Vehicle Routing Problem. The initial objective is to minimize traveling costs. However, in certain contexts, other objectives arise. In home services, a relevant objective is to offer equity between drivers. Equity can be both in terms of profits, traveling distance or duration and difficulty of work. The aim of this thesis is to introduce equity criteria and propose efficient solution algorithms in mobility systems for home services.

Tactical staff planning problems

Optimized management of heterogeneous transportation in industry 4.0

  • Ph.D. Student: A. Benzoni
  • Grant: CIFRE industrial research agreement with STMicroelectronics
  • Start: June 2019
  • End: May 2023
  • Advisor: C. Yugma
  • Description

    Semiconductor wafer fabrication includes one of the most complex manufacturing process. The complexity of wafer production environment results in a very complex transport system. The long lots cycle time, re-entrant wafer flows, several hundreds machines linked to up 700 process steps and different types of scheduling problems (batch, secondary resources, . . .) make transport of lots a very challenging task. While in the most 200-mm wafer fabs material handling is carried manually by operators or with semi-automated transport, wafers and reticles are transported in fully automated manner in modern 300-mm wafer fabs, using Automated Material Handling System (AMHS). In the last years, due to transition from High Volume-Low Mix to Low Volume-High Mix wafer fab, also 200mm sites, that has originally been conceived for manual handling,have started a transport automation process. This thesis focuses on the modernization of the trnsport system of the 200 mm siste of STMicroelectronics in Rousset. The objectives of the thesis can be divided into two major axes: the first part is concentrated on the consolidation of the already implemented AMHS in the photolithography area, by optimizing vehicle and storage management.
    The second focus is on the extension of the AMHS to other workshops, by considering the cohabitation of different transport systems.

Vehicle routing and energy management for a fleet of hybrid/autonomous vehicles

  • Ph.D. Student: M. Trotta
  • Grant: LabEx IMobS3 & FEDER – Auvergne
  • Start: April 2019
  • End: January 2023
  • Advisor: D. Feillet, A. Quilliot (LIMOS)
  • Description

    The use of self-driving vehicles has several aims, including reducing the number of accidents due to human error, reducing traffic in urban areas and improving transport efficiency. The latter is important in order to reduce the environmental impact of transport. Another way to reduce the environmental impact and the concentration of pollutant emissions in cities is to use electric or hybrid vehicles. The aim of this thesis is to propose efficient models and algorithms for the routing and scheduling of autonomous vehicles (electric or hybrid) and for the management of the energy they need to carry out activities related to the transport of goods and people.

Data analytics for intelligent maintenance planning

  • Ph.D. Student: Alexandre MORITZ
  • Grant: CIFRE industrial research agreement with STMicroelectronics
  • Start: March 2019
  • End: January 2023
  • Advisor: S. Dauzère-Pérès
  • Description

    The thesis aims to develop a tool for intelligent maintenance planning (Operational Research) and to propose a methodology to gradually integrate the state of health of equipment (Big Data). Starting from the analysis of field practices for the diagnosis of the state of health or the level of performance of equipment on the one hand and the review of the processes and decision-making tools used in production on the other hand, the objective of the thesis is to propose and develop a methodological approach allowing the progressive integration of the various sources of information available in the planning of the activities of maintenance and the decisions of scheduling of the flows produced.

Theses defended before 2023

Decision support tools for the analysis and optimization of an automated transport and storage system

  • Ph.D. Student: L. Aresi
  • Grant: CIFRE industrial research agreement with STMicroelectronics
  • Start: January 2019
  • End: September 2022
  • Advisor: C. Yugma, S. Dauzère-Pérès
  • Description

    In modern production units such as Crolles 300, unitary storage locations (bins) are placed on the ceiling above the machines to guarantee quick access to store and retrieve lots between two production or measurement operations. . Because the number of bins is limited (currently around 3,500) and cannot cover the total storage capacity required, collective stockers with a large storage capacity (several hundred lots each) are necessary. Collective stockers have a much longer retrieval time than bins. The transport system may transport a lot:

    • directly from one machine to another (not often possible because the destination machine must be available),
    • from one machine to a bin or a collective stocker or
    • from a bin or a collective stocker to a machine.

    In the AMHS (Automated Material Handling System) management system, each machine is associated with a group of bins, where lots awaiting machine availability will be placed. If no bin is available in the group, the lot is taken to a collective stocker. The use of bins helps to greatly reduce the transport time of a lot to a machine compared to collective stockers. However, it is essential to optimize the allocation of bins to machines and to define lot storage management policies by linking them with transport management policies. Thus, the analysis of the impact of the bin allocation on the performance of an AMHS, and the optimization of this allocation and storage management are the major objectives of this thesis.

Modeling and solving capacitated lot-sizing problems with setup times and inventory constraints

  • Ph.D. Student: M. Charles
  • Grant: CIFRE industrial research agreement with Decision Brain
  • Start: October 2018
  • Defense: December 2022
  • Advisor: S. Dauzère-Pérès
  • Description

Modélisation et minimisation du risque en fabrication microélectronique

Product Qualification Management

  • Ph.D. Student: A. Perraudat
  • Grant: CIFRE industrial research agreement with STMicroelectronics
  • Start: December 2017
  • Defense: January 2021
  • Advisor: S. Dauzère-Pérès
  • Description

    In semiconductor manufacturing, each product requires hundreds of operations before completion. An operation must be qualified, in terms of quality, yield and defectivity, which is often time-consuming and expensive, on a machine if it must run the operation. Qualification management is then a critical component of operations management as qualifications are required to configure the production capacity of work centers, shared between all operations, to manufacture products. In the thesis, we are interested in optimizing qualification management both at the operational and tactical levels: We want to optimize the workload balance of the machines, meet the product demand and make the production system more robust against demand uncertainty at the lowest possible cost. New optimization models and solution approaches are proposed and compared.

  • Associated publication(s):  https://doi.org/10.1016/j.omega.2021.102537

Novel optimization approaches for global fab scheduling

  • Ph.D. Student: F. Barhebwa-Mushamuka
  • Grant: European Productive 4.0 Project
  • Start: October 2017
  • Defense: November 2020
  • Advisor: C. Yugma, S. Dauzère-Pérès
  • Description

    This thesis addresses the short-term decision-making problem of Global Scheduling in Semiconductor Manufacturing with emphasis on Consistency between Global and Local Decisions, i.e. Global Strategies and Objectives need to be followed locally, with some flexibility. The main objectives of the thesis are: Balancing Work-In-Process, Maximizing Throughput, Controlling Cycle Times, Minimizing Cycle Times and Minimizing the Variability of Finished Products Mix.

Computing Multimodal Journeys : A Distributed Approach

  • Ph.D. Student: S. Shorten
  • Grant: Industrial research agreement with Cityway
  • Start: October 2017
  • Defense: December 2022
  • Advisor: D. Feillet
  • Description

Production planning for industrial symbiosis

  • Ph.D. Student: E. Suzanne
  • Grant: Bourse Mines
  • Start: September 2017
  • Defense: April 2021
  • Advisor: N. Absi, V. Borodin
  • Description

Planification optimisée des flux dans une chaîne logistique complexe

  • Ph.D. Student: S. Beraudy
  • Grant: European project Productive4.0
  • Start: May 2017
  • Defense: September 2020
  • Advisor: S. Dauzère-Pérès, N. Absi
  • Description

    Within the European project PRODUCTIVE 4.0, the objective of this thesis is to develop and validate advanced approaches to support the management of complex supply chains such as the ones of semiconductor manufacturers and their customers. Supported by a literature review and in cooperation with the industrial partners of the project,  one or several planning models covering the various objectives and constraints of the problem will be proposed.  Approaches to solve the proposed models will then have to be developed and validated on industrial data.

Heterogranular multivariate analytics for detecting and controlling the root causes of the mismatching machines in semiconductor manufacturing

  • Ph.D. Student: A. Chouichi
  • Grant: CIFRE industrial research agreement with STMicroelectronics
  • Start: December 2016
  • Defense: January 2020
  • Advisors: C. Yugma, J. Blue (NTU)
  • Description

    In modern semiconductor fabrication, process becomes more complex and variability propagates through the long production path. Chamber mismatching may be harmless at certain moment but definitely a killer in the final yield. It is therefore necessary to identify the set of all parameters of interest (SPC, FDC, maintenance logs, etc.), to ensure the measurement integrity and the chamber alignment. The thesis aims to develop a methodological approach based on the exploitation of data currently available as well as the development of a decision support tool for the correction of these anomalies.

Development of Global Control Plans in Semiconductor Fabrication

  • Ph.D. Student: W. T. Yang
  • Grant: Bourse Mines
  • Start: November 2016
  • Defense: January 2020
  • Advisors: A. Roussy, J. Blue, M. Reis (EPFL)
  • Description

    : The Run-to-Run (R2R) models within the Advanced Process Control (APC) framework, though well established, are always built based on physics and chemistry knowledge. Since the R2R models can be put in effect only when quality information is measured, they become less timely compared to the real time equipment and production plan. It has motivated the main concept of this thesis where real time and predictive equipment condition shall be adaptively considered in the R2R models such that the process runs are to be fine-tuned for better quality control, more effective production flow, and more efficient equipment utilization.

Modélisation dynamique de la capacité de fabrication pour optimiser la planification de la chaîne logistique

  • Ph.D. Student: Q. Christ
  • Grant: CIFRE industrial research agreement with STMicroelectronics
  • Start: October 2016
  • Advisor: S. Dauzère-Pérès, N. Absi
  • Defense: January 2020
  • Description

Defended theses since 2008

Advanced equipment behavior modeling for process control

  • Ph.D. Student: H. Rostami
  • Grant: ANR
  • Start: November 2015
  • Defense: December 2018
  • Advisors: C. Yugma, J. Blue
  • Description

     Characterize the equipment behavior towards failures in the space of selected moving statistics features and develop an EHI prognosis method for the evaluation of equipment condition after a wafer being processed to identify potential root causes. Furthermore, provide the link of equipment behavior and product quality to Task 3-5 for better product sampling, production scheduling and PM planning.

Models for vehicle routing problems in urban areas

  • Ph.D. Student: H. Ben Ticha
  • Grant: LabEx IMobS3 & FEDER – Auvergne
  • Start: November 2014
  • Defense: November 2017
  • Advisors: N. Absi, D. Feillet, A. Quillot (LIMOS)
  • Description

    The objective of this thesis is to propose new models for Vehicle Routing Problems in urban area that take into account the different criteria involved. Based on these models, we shall develop new solution approaches.

Multimodal, multi-criteria shortest path optimization in road and public transportation networks

  • Ph.D. Student: A. Iglesias
  • Grant: Industrial research agreement with Cityway
  • Start: April 2013
  • Defense: November 2017
  • Advisor: D. Feillet
  • Description

    The objective of this thesis is to develop and improve an existing trip planner engine in collaboration with Cityway. This work consits of three phases: analysis of exising models and algorithms for multimodal routing; multi-criteria variants with pareto optimality considerations; robustness as a new criteria and integration of trip uncertainty.

Consistency of global and local decisions in semiconductor manufacturing process

Characterization of high density plasma etching processes to define advances memories and development of control methodologies of their variability

  • Ph.D. Student: M. Rizquez
  • Grant: Project MAGE
  • Start: May 2013
  • Defense: May 2016
  • Advisor: A. Roussy

Development of innovative approaches for scheduling in a complex semiconductor manufacturing environment

Statistical Learning on Circular Domains for Advanced Process Control in Microelectronics

  • Ph.D. Student: E. Padonou
  • Grant: ARMINES
  • Start: June 2013
  • Defense: May 2016
  • Advisors: O. Roustant (INSA), J. Blue

Robust optimization in the integrated planning of railway resources

  • Ph.D. Student: A. Zehrouni
  • Grant: ARMINES
  • Start: March 2013
  • Defense: May 2016
  • Advisors: D. Feillet, X. Delorme, F. Grimaud, O. Guyon (SNCF)
  • Description

    This thesis is carried out in cooperation with SNCF. The aim is to provide an integrated planning of railway resources (i.e., infrastructure, rolling stock and drivers) in order to minimize delays and costs of the railway operator and passengers.

A mathematical model to optimize regular criteria for the multiprocessor job-shop scheduling problem

  • Ph.D. Student:A. A. García León
  • Grant: Programa de Movilidad Doctoral Colombia hacia Francia Metas 2019, ASCUN, Colfuturo, Colombian Ministery of Education, French Embassy in Colombia, Universidad de Ibagué
  • Start: February 2013
  • Defense: May 2016
  • Advisor: S. Dauzère-Pérès
  • Description

    The multiprocessor job-shop scheduling problem MJSSP is an extension of the classical job-shop scheduling problem JSSP in which each operation must be processed on one machine chosen among a finite subset of candidate machines. The objective of JSSP is to determine a sequence of operations for each machine so that the resulting schedule optimizes regular criteria. Most studies aim to minimize the makespan (the maximum completion time of all jobs in a workshop). However, other indicators such as total flow time, total weighted flow time, total weighted tardiness and customer service (which are also regular criteria) are more relevant in practice and widely used to manage operations effectively as they represent customer service, productivity and competitiveness more accurately. The objective of this thesis is to develop a mathematical model for MJSSP that is able to represent those criteria.

Consistency of scheduling decisions in semiconductor manufacturing

  • Ph.D. Student: A. Bitar
  • Grant: Ministry of Economy, Finance and Industry (MINEFI)
  • Start: October 2012
  • Defense: December 2015
  • Advisors: S. Dauzère-Pérès, C. Yugma
  • Description

    The thesis is about scheduling problems with parallel machines and specific constraints from semi-conductor manufacturing. We apply exact solution approaches using mathematical programming and advanced algorithms, as well as meta-heuristics to solve these complex industrial problems. Some problems include the multi-criteria aspect. This thesis also aims to give a guideline on how to choose the right objective function for local optimization of different areas of a semi-conductor fab in order to provide a good global result.

Capacity modeling and optimization of SOI microelectronic production units

  • Ph.D. Student: M. Rowshannahad
  • Grant: CIFRE industrial research agreement with SOITEC
  • Start: April 2012
  • Defense: May 2015
  • Advisors: S. Dauzère-Pérès, N. Absi
  • Description

    In this research project, we first model and optimize the production capacity of a microelectronic fabrication unit through qualification management. Based on these achievements, we shall propose an integrated production planning model.

Agile models for the agile factory of the future

  • Ph.D. Student: A. Ben Amira
  • Grant: CIFRE industrial research agreement with STMicroelectronics
  • Start: December 2011
  • Defense: October 2015
  • Advisors: S. Dauzère-Pérès, C. Yugma
  • Description

    This thesis aims at proposing and validating agile models adapting to changes of business processes in semiconductor manufacturing. The goals are to render the production system more agile and flexible and to ensure an integrated decision making at different levels.

Optimization of transportation and storage decisions in the automated material handling system (AMHS) of an advanced semiconductor manufacturing facility

  • Ph.D. Student: A. Ben Chaabane
  • Grant: CIFRE industrial research agreement with STMicroelectronics
  • Start: May 2011
  • Defense: June 2015
  • Advisors: S. Dauzère-Pérès, C. Yugma
  • This thesis aims at proposing and validating novel strategies and optimization methods for managing the Automated Material Handling System (AMHS) in semiconductor manufacturing facilities (fabs). The goal is to ensure optimized tactical and operational transportation and storage decisions.

Wafer at risk reduction in semiconductor manufacturing

  • Ph.D. Student: G. L. Rodriguez Verjan
  • Grant: CIFRE industrial research agreement with STMicroelectronics
  • Start: February 2011
  • Defense: July 2014
  • Advisors: S. Dauzère-Pérès, C. Yugma
  • Description

    This thesis aims at proposing and validating novel strategies for wafer at risk reduction in semiconductor manufacturing facilities.

Integrated optimization of production planning and scheduling in a supply chain

  • Ph.D. Student: E. D. Gomez Urrutica
  • Grant: Luxemburgish government
  • Start: November 2010
  • Defense: June 2014
  • Advisors: S. Dauzère-Pérès, R. Aggoune (Université du Luxembourg )
  • Description

    This thesis studies strategies for integrating production planning and scheduling in supply chain management. In practice, decisions concerning these two planning levels are taken in a hierarchical order. Consequently the production plan may generate unfeasibility problems at the scheduling level. The goal of the thesis is to design an integrated approach to integrate production planning and scheduling decisions, in order to guarantee feasible and optimal solutions for complex manufacturing systems and supply chains.

Modeling, characterizing and optimizing the annealing process for manufacturing CIGS solar cells

  • Ph.D. Student: F. Oliva
  • Grant: CIFRE industrial research agreement with NEXCIS
  • Defense: April 2014
  • Advisors: P. Collot, A. Roussy
  • Description

    This thesis aims at modeling and optimizing the annealing step in a two-step process for manufacturing Cu(In1-xGax)(Se1-ySy)2 solar cells (CIGS). The goal is to study the annealing process of electrodeposited metallic precursors and to observe the influence of crucial parameters on absorber characteristics.

A new system for mutualized distribution of goods in cities

  • Ph.D. Student: D. Cattaruzza
  • Grant: Project MODUM
  • Defense: March 2014
  • Advisors: D. Feillet, N. Absi
  • The thesis is part of the MODUM project. The goal of MODUM is to model and quantify the expected gain of a system of urban freight distribution that concerns economical, environmental and social facets. The thesis focus on operational aspects of MODUM and studies vehicle routing problems that arise in the project context and in city logistics in general.

Operations management at container terminals using advanced information technologies

  • Ph.D. Student: E. Zehendner
  • Grant: Ministry of Economy, Finance and Industry (MINEFI)
  • Defense: October 2013
  • Advisor: D. Feillet
  • Description

    This thesis uses information, provided by advanced information technologies, to improve operations at a container terminal. It specially focuses on the allocation of straddle carriers (internal material handling resources) and the organization of the storage area.

Modéliser, simuler et optimiser l’introduction de nouvelles technologies dans une chaîne logistique complète

  • Ph.D. Student: L. Haouari
  • Grant: Project GEOCOLIS
  • Defense: December 2012
  • Advisors: D. Feillet, N. Absi

Modélisation et optimisation intégrées de l’utilisation de l’infrastructure et des engins ferroviaires

Implementing dynamic control plans in semiconductor manufacturing

  • Ph.D. Student: J. Nduhura
  • Grant: CIFRE industrial research agreement with STMicroelectronics
  • Defense: October 2012
  • Advisors: S. Dauzère-Pérès, S. Bassetto, C. Yugma

Ordonnancement et contrôle avancé des procédés pour la fabrication de semi-conducteurs

  • Ph.D. Student: A. Obeid
  • Grant: Ministry of Economy, Finance and Industry (MINEFI)
  • Defense: March 2012
  • Advisors: S. Dauzère-Pérès, A. Ferreira, C. Yugma

Etude, modélisation et intégration de nouveaux concepts de maintenance et de gestion de configuration d’un aéronef à l’aide de technologies RFID embarquées

  • Ph.D. Student: C. Jimenez
  • Grant: CIFRE industrial research agreement with Eurocopter
  • Defense: March 2012
  • Advisor: S. Dauzère-Pérès

Technologies Modélisation et aide à la décision pour l’introduction de technologies communicantes en milieu hospitalier

  • Ph.D. Student: S. Housseman
  • Grant: Project MISTRALS
  • Defense: April 2011
  • Advisors: D. Feillet, X. Xie, N. Absi

Simulation et optimisation du transport automatisé dans la fabrication de semi-conducteurs

  • Ph.D. Student: J.-E. Kiba
  • Grant: CIFRE industrial research agreement with STMicroelectronics
  • Defense: November 2010
  • Advisors: S. Dauzère-Pérès, C. Yugma

Ordonnancement des systèmes flexibles de production sous contraintes de disponibilités des ressources

Modélisation et aide à la décision pour l’introduction des technologies RFID dans les chaînes logistiques

Etude et élaboration d’un contrôle avancée de la production : run to run Grille pour les technologies inférieures à 65 nm

  • Ph.D. Student: N. Jedidi
  • Grant: CIFRE industrial research agreement with STMicroelectronics
  • Defense: October 2009
  • Advisors: S. Dauzère-Pérès, A. Roussy

Replanification de transport après perturbations

Modélisation et optimisation du management de recettes et de qualifications d’équipements

  • Ph.D. Student: C. Johnzén
  • Grant: CIFRE industrial research agreement with STMicroelectronics
  • Defense: April 2009
  • Advisor: S. Dauzère-Pérès

Etude et validation de boucles d’asservissement permettant le contrôle des procédés de fabrication en microélectronique

  • Ph.D. Student: D. Belharet
  • Grant: Project ROUSSET 2003-2008
  • Defense: February 2009
  • Advisors: P. Collot, A. Roussy

Techniques hybrides de recherche exacte et approchée : application à des problèmes de transport

Planification optimisée des opérations dans les établissements de maintenance du matériel roulant de la SNCF

  • Ph.D. Student: F. Ramond
  • Grant: CIFRE industrial research agreement with SNCF
  • Defense: May 2008
    Advisor: S. Dauzère-Pérès

Approche intégrée en planification et ordonnancment de la production

Transport automatisé dans les systèmes de fabrication de semi-conducteurs : Nouvelles approches de gestion tactique et opérationnelle

  • Ph.D. Student: J.R. Montoya Torres
  • Grant: CIFRE industrial research agreement with STMicroelectronics
  • Defense: November 2005
  • Advisors: J.P. Campagne (Ecole Centrale de Lyon), S. Dauzère-Pérès, H. Marian