Batching issues tend to have a profound influence on the flow characteristics of a manufacturing plant, and substantial gains can be made by properly understanding underlying principles. Almost all kinds of automotive operations are associated with large batching operations. The following are the commonly found large batching operations in the automotive manufacturing units:
- Forging – Forging Presses
- Casting – Core making and Moulding Machines
- Plastic Components – Plastic Moulding Machines
- Sheet Metal – Metal Presses
Conventional operational thinking is based on maximizing the utilization of these batching machines. But one of the major issues with large batch sizes needed for maximization of capacity utilization is that it increases the overall production lead time and leads to poor on-time delivery to customers. This is because executing large batches results in ‘capacity stealing’ – i.e. the capacity is utilized for the production of items that are not needed in the immediate time horizon at the cost of immediately required items and/or items required for confirmed orders.
Large batches at batching machines can also create ‘Capacity Waves’ at downstream operations. This is largely due to the fact that the space available between the large batching machines and downstream operating machines/lines is limited and downstream operations are forced to ‘suck’ the inventory created by the large batching machines.
Figure 1: “Peak and Lean Production Periods”
The situation becomes worse in the cases where the batching machine is feeding multiple manufacturing lines downstream. Such a phenomenon creates ‘lean’ and ‘peak’ production periods across lines. ‘Lean Periods’ are essentially near starvation of the downstream processes and may lead to the idling of labour, whereas during ‘Peak Periods’, the downstream processes are flooded, often leading to overtime situations. A figurative representation is presented in figure 1.
The conventional way of operation which essentially focuses on utilization, with the above-mentioned consequences, makes them appear as constraining machines, and planners plan the entire production considering these machines as constraints. On the contrary, in most of the plants, it has been our experience that the batching machines have never emerged as internal constraints in the manufacturing setup. These batching machines tend to have sufficient excess capacity. The deluge of inventory lying past these machines clearly points out that these machines are not constraints to the flow in such situations. The real constraints have largely been found to be in the downstream operations! For instance, in the case of sheet metal operations, the constraint is mostly found in the welding and finishing operations, whereas in castings, it is found in fettling and finishing operations.
As mentioned earlier, increased batch sizes will decrease throughput and lead to poor on-time delivery to customers. Moreover, increased batch size also affects work-in-process inventory levels, manufacturing lead time, local and global safety time issues, and finished goods stock levels by increasing them. Therefore, since batching on a non-constraint at the plant is undesirable – ideally, batch sizes should be reduced to levels required for consumption in downstream processes or for orders in the immediate horizon. But, on the other hand, smaller batches increase the number of setups which in turn can totally erode the protective capacity of the non-constraint machine converting it to a constraint machine. This can also then result in poor on-time performance.
This phenomenon of a batching process can be depicted by an ‘Inverted U Curve’ where both extremes are non-optimal points and should be avoided. The optimal capacity and high ‘On-Time Performance’ occurs somewhere in the middle of the U Curve. This means that large batches with an eye on utilization or very small batches based on immediate consumption can both be detrimental to the plant. The optimal point for different machines with different capacities needs to be identified through an iterative process, but it will be found between these extremes (Refer figure2).
In such scenarios of batching operations, it is also essential to decouple the batching machine from the downstream operations. This is to allow these machines to have the flexibility to be out of sync with the rest of the operations downstream without creating disruptions of in flow in either. Placing buffers immediately after the machines will result in a capacity release from the existing batching machines and help in moving close to the optimal capacity. The batching machines will now respond to the trigger received from their buffer consumption, making it possible for these machines to take larger batches. But since the downstream activities now consume from this buffer, the cycles of starvation and flooding are also averted. These downstream activities will act as an independent system with its WIP being controlled by a trigger from the FG inventory consumption and ensuring on-time delivery to the consumer.
Figure 2: “Inverted U Curve” of a batching machine
It may sometimes be observed that capacity release obtained by reducing the batch size (i.e. stopping excess production of only one type of component) may not be enough to offset an extra set-up. This is a phenomenon where the ‘Inverted U Curve’ has the points A, B, and C very close to each other. This implies that the spare capacity available at the batching machine is very less, and immediate improvement techniques based on SMED has to be deployed.
The scheduling can be done on the basis of the bottlenecks and the demand can act as the input for the bottlenecks. This will eliminate the unnecessary inventory buildup downstream as well as need for decoupling. Can the batch sizing process be optimized by combining or splitting different orders?
Hi Parikshit,
The type of scheduling depends on the tolerance time of customer. For instance if the tolerance time of customer is less than the supply lead time, then we will have to decouple downstream by keeping stocks so that the customer gets the material as and when he requires immediately. If the tolerance time is more than the supply lead time then stocks may not be required and the orders placed can be scheduled on the bottleneck and supplied on the agreeable lead time.
However, the article discusses scenarios where a large non bottleneck batching machine upstream may create a possibility of capacity stealing as well capacity starvation of downstream operations creating under load and overload situations.
As long as clubbing and splitting of orders do not affect the individual due dates of orders the same can be looked into, however as clubbing or splitting increases one can move to either side of the inverted U curve which may affect due dates.