An alternative to implementing multiple modules like flow and discrete for manufacturing, the process-mix can be handled using Discrete Manufacturing module alone, resulting in lowering of implementation and support cost.
Manufacturing companies usually have a mix of discrete and flow type manufacturing processes. An alternative to implementing multiple modules like flow and discrete for manufacturing, which may make the implementation complex, the process-mix can be handled to a large extent using Discrete Manufacturing module alone, resulting in lowering of implementation and support cost.
The shop floor operations of any manufacturing entity are normally a mixed bag of discrete, flow, and processing at the outside vendor (subcontracting) etc. Implementing multiple modules of Oracle (Flow Manufacturing, Discrete Manufacturing) to handle the operations will make the implementation very complex and costly. Also the data and setup requirements for modules like flow manufacturing are quite demanding. This paper provides an alternative option to model the mixed mode manufacturing setup (combination of OSP, in-house discrete, in-house cell manufacturing) using conventional WIP and BOM modules, and simplify the setup by not having to use Flow or Network routings. This will reduce the implementation and maintenance cost but will give fairly accurate results.
Mix mode manufacturing — Concept and Relevance:
The key to success in manufacturing today is it’s agility — the ability to precisely anticipate and swiftly respond to fluctuating market demands. As the supply chain becomes more demand-oriented, manufacturers are under pressure to bring products to market in the shortest span of time, at the right price, and adhering to the customer’s specifications. Moreover, during hard economic times, organizations scramble to find business, any business to help ride out of storm is welcome.
To achieve this, more and more companies are pursuing mixed-mode manufacturing and employing different production strategies for different product lines. One will therefore mostly find that the shop operations are a combination of different manufacturing techniques like discrete manufacturing, repetitive manufacturing, outside processing, single piece flow manufacturing, etc.Pit Falls
A lot of ERP packages have separate modules for discrete, flow, repetitive, etc. In order to map the shop floor operations ideally, one many need to implement a plethora of modules. This makes the implementation costly due to high licensing, consulting and maintenance cost, and training cost. Also, the implementation timelines will go up and benefits of accrual of the implementation gets delayed. The master data requirements and the accuracy of data that these multiple modules demand are also very high and it is quite difficult in lot of cases to really generate that data with the existing processes and systems.
One has to really do due diligence on whether we really need that level of accuracy in terms of replicating the exact shop floor scene in Oracle. Can we sacrifice some accuracy in terms of replicating the shop floor model exactly in our ERP system, but reap the benefits like a simpler and cost effective implementation with a non complex system to handle during its life cycle. Below section explains how different manufacturing processes are handled using discrete manufacturing module.
Mixed-mode Manufacturing: How to implement using Discrete Manufacturing
Choosing a software package that handles multiple manufacturing processes is a no brainer for those who are currently involved in mixed mode manufacturing, but it is also a good idea for those who don’t because you never know what the future holds. When times are bad and you’re scrambling for work, your ability to react to the market and provide good service has a lot to do with the tools your people have at their disposal.
The below section explains how Mixed mode manufacturing can be implemented using Discrete Manufacturing by using some workarounds in the way the operations are being modeled.
Single Piece Flow:
In this kind of manufacturing process, pieces flow in single piece for each operation rather than lots. Once the first operation is complete on that piece, it has to be moved on for the second operation and the next piece is moved on for first operation.
The diagram above depicts a single piece flow scenario. Every piece move from Turning to Welding and then to Assembly. Once first piece completes Turning operation and moves to Welding, the second piece starts the Turning operation, and so on.
Manufacturing companies usually have a mix of discrete and flow type manufacturing processes. An alternative to implementing multiple modules like flow and discrete for manufacturing, the process-mix can be handled using discrete manufacturing module alone, resulting in lowering of implementation and support cost.Let’s take an example of single-piece flow with a sample routing for item XXX.
If we load a job of 50 for the assembly XXX, it will be completed on the shop floor as shown below:
Thus, even total lead time of producing 1 piece is 30 min., a job of 50 quantity doesn’t take 50*30 i.e. 1500 minutes. But it takes only 520 minutes because of single-piece flow.
If each of this operation is modeled as individual operation in oracle routing and a discrete job is loaded for 50 quantity, Oracle discrete job jcheduling will schedule this job as:
Thus oracle discrete job scheduling will give us 1500 minutes as total job lead time (Against an actual of 520 minutes) if we model each operation as item based operation in routing.
Mapping of actual physical operations directly in discrete manufacturing may not work. Discrete manufacturing by definition assumes that the entire job quantity is loaded simultaneously on one operation and moves to next operation as a lot. But that is not the case in single piece flow operations.
In actuality, the time required to complete the job in single piece environment is very close to the quantity multiplied by the ‘drum-beat’ rate of the sequence of operations.
Hence the single-piece flow has to be dealt by considering all three operations as one single cell / operation and using the drum beat of 10 minutes as the resource usage.
Batch Operations: In this kind of manufacturing process, pieces flow in batches where processing of all the pieces is started at the same time, and finished at the same time. A batch production is started when there is enough quantity (for example, minimum quantity needed to start an oven for energy consumption reason) available to start the operation.
A batch type of operation takes same time for completion for job quantities from 0 to maximum quantity that can be processed in a batch. For e.g. drying oven can have 50 components at a time in oven and takes 3 hours to complete, then even if 1 piece is loaded in the oven or 50 pieces are loaded in the oven, it will take 3 hours to complete.
In the above example, Operation 30 is batch operation and after Operation 20 all pieces wait till the batch quantity is available to be loaded in Operation 30.
Let’s take an example with a sample routing with all batch operations for item YYY.
If we load a job of 50 for the assembly YYY, it will be completed on the shop floor as shown below:
Outside Processing Operations (OSP): In this kind of manufacturing process, material is sent to outside processing subcontractor in lot. Material is sent to outside vendors for processing, and when there are multiple OSP vendors, material sometimes go from one vendor to other vendor directly and return after completion of all outside operations.
There is an element of documentation time before dispatch, transportation time and wait (Queue) time at sub-contractor’s place and actual operation time in outside processing operation involved.
Let’s take for example, Material is sent to OSP vendor against a job with following details:
Quantity = 50 Dispatch Documentation preparation and execution = 0.5 day Transportation time from plant to OSP = 0.5 day Operation time per piece = 10 min Transportation time back to plant = 0.5 day
Different operations and resources capturing all these lead times will make the routing more complex. In order to reduce the complexity, it can be handled by combining all item based lead times on one resource and combining all lot based lead on another resources of the OSP operation.
As shown above, lot based operation time is 1.5 days irrespective of WIP job quantity. Item based resource time is calculated by multiplying usage and WIP job quantity
Lot based operation time = Dispatch documentation + Transportation from plant + Transportation to plant Item based operation time = Operation time per piece
Mixed Mode: Most of the manufacturing industries have combination of different modes of manufacturing. Below is the depicted picture of a routing with mixed modes of manufacturing.
Operation 10 is a discrete process followed by Operation 20 which is a single-piece flow (cell with all three operations turning, welding and assembly clubbed) and then operation 30 is a single-piece operation and operation 40 represents batch operation and finally operation 50 is an OSP operation.
Even though this simplifies the modeling of shop operations, there are a few drawbacks:
1. In certain cases, we may need to define separate resources for costing purposes since this type of modeling may not handle all the costing requirements.
2. In case of single piece flow normally quality is checked at the end of all the operations. In case if quality check / disposition take place within the operations, this model may not yield correct results.
3. The model for single piece flow will not work in case if the line / part of the line which has single piece flow is not balanced.
Mix mode manufacturing is here to stay, and companies need to figure out the best way to model the operations in the ERP packages that are implemented. Rather than going in for multiple modules to handle different types of manufacturing practices, they may go for conventional modules only like discrete manufacturing, but with a bit of workaround to model the operations fairly well and achieve good results. This will make the implementation and maintenance of system far simpler and cost effective.
About the authors:
Yogesh is a lead consultant in the Enterprise Solutions group of Infosys Technologies Limited.
Shantanu is a senior consultant in the Enterprise Solutions group of Infosys Technologies Limited.
Vamsi is a consultant in the Enterprise Solutions group of Infosys Technologies Limited.