Great news! I have managed to publish my PhD Thesis developments in yet another prestigious scientific journal – The Journal of Cleaner production!
Our goal: Make manufacturing system and machines more sustainable, that is more energy-efficient, but without sacrificing their productivity
What is it all about?
Long story short, this time me and my colleagues Dr. Bianchi and Prof. Tolio have dealt with the optimization of machining cycle in machine tools.
I think the abstract of our work speak best for itself:
The study addresses sustainable manufacturing, focusing of efficiency management in machine tools. A novel approach to energy modelling and machining cycle optimization is introduced. It uses a hierarchical optimization approach, splitting machining cycles into smaller, independent tasks. The analysis starts at the level of workpiece design features, for which a variety of state-of-the-art energy models can be used to link process and machine tool parameters to the resulting energy use. The Pareto front resulting from local optimization of each feature is represented by functions relating execution time and the corresponding minimal energy consumption. This reduced information is passed to the higher levels of the task tree, up to processing units (in multi-spindle machines) and machine level. Propagation is performed through local optimizations of serial or parallel execution of the underlining tasks: under the specified assumptions, the merging process does not affect the global optimality of the solution. This approach strongly reduces problem dimension, allowing to postpone at machine level the optimal selection of all processing parameters. The proposed approach is therefore applicable to system wide analysis of multi-machine manufacturing systems, where machine and system level decisions are tightly coupled. A case study of an industrial flexible transfer machine, on which the developed methodology was applied, is analysed and discussed.
Essentially, what we proposed is a method for modelling of manufacturing process in a machine with a hierarchical tree of operations and task, just like the one here:
The second step is to perform optimization, by optimizing firstly all the tasks in the tree, starting from the simples ones (geometrical feature of the workpiece) until a grouping tasks of higher levels (like overall task of a processing unit of finally: of the whole machine). At each level we perform multi-objective optimization aiming at reduction of both processing time and energy use. What is distinct of our approach the final result is not just one value of cycle-time and energy, but a function of one variable depending on the other one. This way we can dynamically reprogram machine according to our current needs and also state of other machine is a manufacturing lines, ensuring we are operating in the most efficient regime our machine can provide. This is symbolically represented by our design space (all possible solutions) and the, so-called, Pareto frontier (all optimal solutions):
The orange line show our machine best performance, both in terms of energy and processing time. In the end it is up to machine user to decide which working point on this curve is most suitable, considering his manufacturing context.
Overall, applying our framework to an industrial case study (courtesy of Porta Solutions s.r.l. for making their machine available for testing), we achieved 8% reduction in energy and around 8.3% reduction in processing time!
Where can I get it?
Accepted (not formatted by the journal) version of the manuscript is available for free under this link:
Hierarchical modelling framework for machine tool energy optimization (PDF of accepted manuscript)
For the official journal version, please refer to: https://doi.org/10.1016/j.jclepro.2018.09.030
 J. Wójcicki, G. Bianchi, and T. Tolio, “Hierarchical modelling framework for machine tool energy optimization,” J. Clean. Prod., Sep. 2018.