Alexis Aubry's homepage

Research

Since 2008, I am an associate professor with the CRAN within the research group ISET (Sustainable Systems Engineering).
My research topics concerns the control of production systems when considering perturbations in the context of the Industry of the Future. From a production point of view, the Industry of the Future aims to make possible mass customization of products, large flexibility and agility of manufacturing recipes and resources as well as ability to face a high variability (volume and mix) on product demands. However, this challenges the classical control strategies, which often propose to compute a centralised production schedule (in terms of starting and completion dates of the production operations and allocation of resources) within a stable and certain environment (stable and known demand, pre-defined manufacturing routes, stable features and availability of resources…). The optimal schedule, calculated according to these optimistic assumptions, may be deteriorated during its implementation, given the inevitable deviations in practice from the reference environment: uncertainties and high variability on products (volume and mix, mass customization, production of small series, shorter and shorter delivery time...), uncertainties about production resources (operating times, machine failures…).
In this context, handling production distrurbances in the control of production system is one of the main challlenges to be addressed for making the Industry of the Future a reality. Dealing with this challenge asks the question of the ability to take robust decisions against uncertainty (off-line) or the ability to be flexible (on-line).

This issue falls within the objectives of the french research group GT IMS² of the GDR MACS, and of the IFAC Technical Committees TC 5.1 "Manufacturing Plant Control " and TC 5.2 "Manufacturing Modelling for Management and Control ".

In this context we propose to address the following scientific problems: how to propose a standardized approach for taking into account the perturbations when controlling production systems and then to compute robust or adpatable solutions?

These problems concerns also an industrial need that consists in taking decisions, in a perturbed context, whose risk must be controlled and contained.

In particular, we identify the following scientific questions:
How to model the different perturbations occuring in production systems and their environment?
How to model the problem of production planning and control under perturbations?
How to evaluate the robustness of a production plan or schedule?
What must be a proactive-reactive approach for computing solutions that are able to resist to perturbations and to change when hazards occur?
When to switch between the proactive phase and the reactive phase?

The tools that are used are mainly taken from the Operational Research Theory. Bu we are also exploring an original solution based on Discrete Event Systems models and tool. Some contributions have been proposed and published (see the page Publications for more details).