SUBSCRIBE

Breaking News on Food and Beverage Processing and PackagingWorldUSEurope

News > Processing

Read more breaking news

 

 

Optimizing bakery production: Modified genetic algorithm most ‘promising method’

By Kacey Culliney+

01-Aug-2014
Last updated on 04-Aug-2014 at 14:25 GMT

Bakery production can be optimized best using a modified genetic algorithm, researchers find
Bakery production can be optimized best using a modified genetic algorithm, researchers find

Using a modified genetic algorithm (MGA) to design scheduling plans in an industrial bakery can save costs and resources, scientists say.

Published in the Journal of Expert Systems with Applications, researchers investigated use of three algorithms to develop efficient production scheduling in bakeries - modified genetic algorithm (MGA), ant colony optimization (ANO) and random search procedure.

Findings showed idle time of machines could be reduced by up to 23% and makespan – the required time to complete the given production goal – by as much as 8.6%.

While all three significantly reduced makespan, the MGA approach was touted as the “most promising method” for bakery scheduling in terms of reductions in idle time and makespan.

The MGA approach was the 'most promising' in reducing makespan

“The application of such a mathematical tool for production scheduling instead of using only empirical experience would almost surely lead to a considerable increase of baking companies’ efficiency, saving both production costs and resources in general, since even the worst case results of the applied methods for both cost functions yielded significant benefits compared to the results given by the initial product sequence,” the researchers wrote.

Each algorithm, on average, was able to solve production problems within 15 minutes, but the researchers said this could be optimized even further using speed-up procedures on algorithms or running the program on a high performance PC.

Earlier research: Animal-inspired algorithms

The same team already investigated ACO and particle swarm optimization (PSO) as methods to optimize production in bakery plants. They concluded both methods were comparable in terms of makespan reduction and idle time in machines, nothing therefore it could be interesting to combine the algorithms.

The lead researcher Florian Hecker previously told BakeryandSnacks.com humans could not develop bakeries anywhere near as efficient as those created using algorithms .

“The use of numerical methods just provides new possibilities for creating an enhanced production planning,” he said.

German case study

The researchers used the German bakery sector as the case study for its research, because they said the performance of bakeries was “often sub-optimal”.

“Even though this industry brand with its high diversity of products and time dependent production processes is as if predestined for the application of state-of-the-art scheduling methods. In the German baking industry, the production planning is almost completely based on the practical experience of the responsible employee(s) instead of the usage of mathematical methods,” they wrote.

Testing and analysis was done on a bakery with 26 stages and 40 products.

 

Source: Journal of Expert Systems with Applications
Published online April 2014 ahead of print, doi: 10.1016/j.eswa.2014.03.047
“Applicaion of a modified GA, ACO and a random search procedure to solve the production scheduling of a case study bakery”
Authors: FT. Hecker, M. Stanke, T. Becker and B. Hitzmann

Related products