Design and Implementation of a policy-driven orchestrator for production planning
Elahi, Mahboob (2020)
2020
Master's Programme in Automation Engineering
Tekniikan ja luonnontieteiden tiedekunta - Faculty of Engineering and Natural Sciences
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2020-10-14
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tuni-202009307206
https://urn.fi/URN:NBN:fi:tuni-202009307206
Tiivistelmä
From last few years market trend has become more competitive due to abrupt and unpredictable customer demands. This aggressive and competitive production market environment pushes industrialist to come up with new production policies to meet consumer demand in time by improving their production rate without compromising the product quality and product cost. To achieve this goal manufacturers are using mixed model assembly lines. During MMAL operation many problems are faced by industrialist such as mix-model assembly line balancing and sequencing, equal distribution of workload among workstations to avoid work cell overloading, optimization of cycle time and minimizing of tool changeover time etc.
The aim of this thesis work is to develop a policy-driven orchestrator which accepts different daily production policies to process production lot by considering the MMAL problems like line balancing and sequencing etc. The orchestrator is developed with three different production policies which address MMAL issues related to line balancing and sequencing and optimizing cycle time of production line for available production stations.
The first proposed policy is “nearest free workstation”. It is reference policy and no steps are taken to mitigate MMAL problems. This policy used to get the results of an unbalanced MMAL line which are compared with the results obtained by other two policies to emphasize the importance of optimizing and balancing of MMAL. In reference policy drawing jobs are assigned to the nearest free work cell to workstation 1 in a sequential fashion. In this policy all the cell phone components are drawn on the same workstation to which a pallet enters for production.
The second production policy is “fixed color” policy in which each of the workstation assigned a fixed color out of available three (Red, Green and Blue) colors at orchestration run time to achieve desire cycle time by minimizing the tool changeover time. In this policy cell phone components with specific color are drawn at specific workstation which has same color as cell phone component.
The third production policy is “fixed color and fixed drawing recipe” in which each workstation assigned a fixed color out of available three (Red, Green and Blue) colors and available cell phone drawing recipes such that cycle time constrain is not violated at orchestrator run. For example, all frames of red color are drawn by cell 2, all screens of green color are drawn on cell 3 and so on. In this policy cell phone components are drawn at workstations which have same specifications as of cell phone component.
Furthermore, the developed system provides a user interface where customer places daily production Order. In addition of order placement, customer can see the current status of finished or under processed orders batches in a lot, individual orders of an order batch and subscribes/unsubscribes to production line events. MySQL database is used as storage for production data.
Finally, after implementation, the developed production policies are tested on FASTory Simulator and FASTory line for a test case order received by user from order placement UI. After calculation and analysis of results obtained by these production policies, it concludes that policy 3 is the best policy among the other two production policies with 0% similarity index with other policies for any order with any combination of components and their colors out of available distinct 729 combination.
Furthermore, a close examination is done between policy 2 and 3 for order combinations in which more than one cell phone component has same drawing color. These order combinations transform a production policy to other which affect cycle time and assembly line load balancing constraints.
To conclude, the policy 3 is best to meet the desire MMAL balancing and sequencing and cycle time regard less of any order combination and Policy 2 is good to use when it is sure that there is no order with any kind of overlapping component specifications.
The aim of this thesis work is to develop a policy-driven orchestrator which accepts different daily production policies to process production lot by considering the MMAL problems like line balancing and sequencing etc. The orchestrator is developed with three different production policies which address MMAL issues related to line balancing and sequencing and optimizing cycle time of production line for available production stations.
The first proposed policy is “nearest free workstation”. It is reference policy and no steps are taken to mitigate MMAL problems. This policy used to get the results of an unbalanced MMAL line which are compared with the results obtained by other two policies to emphasize the importance of optimizing and balancing of MMAL. In reference policy drawing jobs are assigned to the nearest free work cell to workstation 1 in a sequential fashion. In this policy all the cell phone components are drawn on the same workstation to which a pallet enters for production.
The second production policy is “fixed color” policy in which each of the workstation assigned a fixed color out of available three (Red, Green and Blue) colors at orchestration run time to achieve desire cycle time by minimizing the tool changeover time. In this policy cell phone components with specific color are drawn at specific workstation which has same color as cell phone component.
The third production policy is “fixed color and fixed drawing recipe” in which each workstation assigned a fixed color out of available three (Red, Green and Blue) colors and available cell phone drawing recipes such that cycle time constrain is not violated at orchestrator run. For example, all frames of red color are drawn by cell 2, all screens of green color are drawn on cell 3 and so on. In this policy cell phone components are drawn at workstations which have same specifications as of cell phone component.
Furthermore, the developed system provides a user interface where customer places daily production Order. In addition of order placement, customer can see the current status of finished or under processed orders batches in a lot, individual orders of an order batch and subscribes/unsubscribes to production line events. MySQL database is used as storage for production data.
Finally, after implementation, the developed production policies are tested on FASTory Simulator and FASTory line for a test case order received by user from order placement UI. After calculation and analysis of results obtained by these production policies, it concludes that policy 3 is the best policy among the other two production policies with 0% similarity index with other policies for any order with any combination of components and their colors out of available distinct 729 combination.
Furthermore, a close examination is done between policy 2 and 3 for order combinations in which more than one cell phone component has same drawing color. These order combinations transform a production policy to other which affect cycle time and assembly line load balancing constraints.
To conclude, the policy 3 is best to meet the desire MMAL balancing and sequencing and cycle time regard less of any order combination and Policy 2 is good to use when it is sure that there is no order with any kind of overlapping component specifications.