Managing T - Plant Material Flow Configuration

Chaos to Order/Clarity to Success

We have had opportunity to work with many clients whose companies had T-Plant Flow. These were Pumps, Motors, LED Lights and Fixtures, Consumer Appliances etc.

The T-Plant is also known as a "many-to-many" plant. The critical feature of a T-plant is that the final products are assembled using a number of component parts and these component parts are common to many different end items (in contrast to an A-plant) The production routings are usually dissimilar for each items but these are in turn used as parts for different final products. The number of end items is larger (much larger) than the number of component parts. This creates the sudden explosion of the PFD to create the T-shape. To illustrate the magnitude of this explosion, consider a case where there are six component parts and each part has four variations, giving a total of 24 different components. The number of possible end products is 4 × 4 × 4 × 4 × 4 × 4 = 4096!. ?

Major industries that have this flow configuration are -

1. Pumps
2. Motors
3. Consumer Durables / Appliances
4. Personal Computers
5. Variants within Automobile Models
6. Furniture
7. Locks

?Challenges of T-Plants:

?T-Plant combine an I or an A Plant for components and a clear V for the final product. The huge variety is created by various combinations (configurations) of components The ‘T’ plants combine the problems of both A and V plants Synchronization problems and materials availability. The challenge of Assembly is to get full-kit of components to start and complete a batch of Finished Goods Assembly. Since, all components are not always available, production planning does what is best in given circumstances and prevent assembly resources from staying idle. FG SKUs which might not be immediately required are produced as, at some point of time, their components are available in full-kit. This action consumes the components which are required for urgent FG SKUs. This creates a dilemma for component availability and utilisation of assembly resources. Should they wait for right components to arrive and produce based on priority or should they work what is available. The other option exercised is to partially assemble the high priority FG with available components. This creates semi-finished goods inventory which clogs the shop-floor.

?

1. Large finished goods and component inventories.
2. Impression on stealing of components
3. Semi assembled FG SKUs waiting for some components to arrive.
4. Poor due date performance (30 to 40 percent of the orders early and 30 to 40 percent of the orders late).
5. Excessive fabrication lead times.
6. Unsatisfactory resource utilization in fabrication.
7. Fabrication and assembly act as separate unsynchronized plants.

?

Experience Sharing of working for clients with T-Plant configuration.

LED Lighting and fixtures

• Starting Reality

We worked for a manufacturer and distributor of LED Lighting and fixtures to projects and institutional customers. They are a division of a large conglomerate which has LED for retail sales also. Only 30% of what they sell is manufactured inhouse. Remaining is contract manufactured from different vendors. These vendors also manufacture for other similar companies in the market. As expected in a T-Plant they had a big variety. Over a period of 24 months they had invoiced 3500 different SKUs. Since the lead time of order execution was long, sales team had got into habit of punching into system customer forecast based order. These were project orders and sales team used their judgement of probability of winning the business and placed the order even before the deal was finalized. In quite a few cases they lost the deal and the order was quietly cancelled in the system. This resulted in CHAOS as cancelling orders to subcontract manufacturers midway when they had already committed their working capital was next to impossible. This led accumulation on made to order inventory which did not have any order. Sales team was tasked with clearing this stock by getting matching order. There was no way to determine genuine demand backed by firm orders. LED has a relatively short lifecycle. A big chunk of non-moving? inventory had components which were now replaced by latest version and hence the customers would not want these in their supplies. Sales team had monthly target and hence billing was bunched in the last week of the month. Operations and supply chain team had a tough time to keep up with billing targets. All review meetings were blame game. They were suffering from the effect of vicious cycle that was operating in their environment.

• Analysis and solution elements

Our first step was to simply the product portfolio and create an environment where the operations delivered orders reliably and with short lead times. With the client team, past 24 month data was analysed for pareto and Head Body Tail SKUs and Customers. (50:40:10 distribution) Armed with different perspectives of the data, we asked the cross-functional team which included heads of sales, marketing, new products development, plant, vendor management, purchase to classify their product basket into MTA, ATO and MTO. This was with the view of defining lead times which were practical. MTA SKUs would always be available subject to maximum order quantity allowed, Assemble to Order would have one week lead-time subject to maximum order quantity allowed, as all the components of the BOM were always available. Made To Order SKUs would have lead-time ranging from 3 to 6 weeks depending on the lead-time of BOM components with longest lead-time, with a condition of minimum order quantity. After a few interactions the cross-functional leadership team came up with a radical consensus 80 SKUs in MTA (2.3% of starting 3500) 80 SKUs in ATO (again 2.3% of 3500) 700 SKUs in MTO (20% of 3500) 2640 SKUs in NO or Refuse to Quote (75.40% of 3500) This was possible after intense debate and looking at their own past data. They realised that their sales team in anxiety to meet their sales target when their lead times were high, was picking up orders which had SKUs which were refused by other major competitors. They decided to educate the customer to pick up SKUs from the newly declared catalogue to experience fast and reliable deliveries. They also decided to review this portfolio every four weeks so that old generation LEDs were replaced with new generation and their inventory always had fresh components

• Emerging Reality

The leadership team stepped up continuous communication with the sales team and their lead generators in the market about the decision to abide by the discipline. The sub-contract manufacturers were also taken into confidence. They were pretty satisfied with this call to be disciplined. Excel based planning engine configured with MTA, ATO and MTO with their Bills of Material was configured. The planning engine had stock based buffers for FG, Components and RM as well as time based buffers for ATO and MTO orders. Planning Engine gave daily visibility to all the stakeholders about the new supply orders backed by firm customer orders that had to be initiated (Demand Driven). They also had the visibility about the priority in which the existing supply orders had to be executed. Sales team got cracking on liquidating accumulated obsolete stock with schemes It took 6 months for the buffers to be built up at all the levels FG, Components and RM. The sub-contract manufacturers had clear visibility of MTA buffer penetrations and firm ATO and MTO orders

?

If you have a T-Plant configuration, would your leadership team be in a position to analyse and drive consensus on such a radical decision? Untamed variety seems to be at the root of CHAOS that T-Plant companies face.

What do you think the impact of these decisions and actions would be on different stake-holders?

Would the sales team start enjoying the reality or would they be continuously looking to add more and more variety?

What would be impact on inventory by these decisions and actions?

?

Contract Manufacturer of small domestic and agricultural pumps and motors for multiple well-known pump brands

• Starting Reality

We worked with pumps and motors manufacturer who worked for multiple well-known pump and motors brands as their contract manufacturer. In some cases, OEM had given their design and drawings and it others, our client had their design and drawings approved by OEM. Working with multiple OEMs they had wide variety of products to cater to. The OEMs used to give monthly projections. Our client used to plan after the receiving the projections and place orders for parts and components to external suppliers and inhouse manufacturing / fabrication shops. Getting full-kits for the assembly to be scheduled was a challenge. It was a daily routine, that first two hours were spent on scrambling for full-kit of material, wasting the assembly capacity leading to overtime by shift extension. Many times the assembly operation had to work with partial kit and leave partially assembled products on the floor waiting for missing part to arrive for final assembly and subsequent steps of testing and painting.

The observed pattern of monthly production and billing was as follows:

• First 10 Days of the month: 20%
• Next 10 Days of the month: 30%
• Final 10 Days of the month: 50%

?A typical hockey stick billing pattern repeating month after month.

?The client was under pressure to increase capacity as they were able to supply only 70 to 80% of the orders made available by OEMs every month. The products had seasonal demand pattern and some months had spike in orders like motors for desert and room coolers demand was concentrated in the months of Jan to May. OEMs dealt with multiple contract manufacturer as each had similar challenges.?

• Analysis and solution elements

We studied the past invoicing patterns and realized that even with multiple OEMs, the client had pareto distribution. Using this insight we classified the SKUs as MTA and ATO. All the components and RM for all the SKUs were buffered and the size of the buffer was modulated based on seasonality for every month.?

We realized that if the full-kit availability of components is ensured then the hockey stick pattern for production and billing could be smoothened to a large extant. Moreover, the assembly operation could work with much larger capacity within the shift timings. Overtime option could be exercised if there was a spike in order. This would normally happen if the competing vendor is not able meet their quota of supply.

We divided the monthly projection into four weekly master production schedule where MTA was supposed to be produced every week and capacity for ATO was reserved for use depending on the short notice from OEMs.

The component buffer decoupled the manufacturing value chain into components production and assembly operation. The component production and upstream purchase orders to vendors and subcontractors were completely demand driven, i.e. as soon as the material is issued to shop-floor, the supply orders for components get generated. Similarly, RM buffer protected inhouse component production from variability from RM supplier. Dispatches from FG Buffer for MTA SKUs triggered replenishment orders to produce MTA SKUs.

An Excel based Demand Driven planning engine consumed data from ERP and was run daily to generate all the supply orders with Black, Red, Yellow and Green colours defining the priority sequence in which the different department were supposed to execute their respective actions.

The assembly operation was not the bottleneck. The inhouse Winding operation had capacity limitation. Closer study revealed that there was scope for increasing output of winding by 100% by proper exploitation steps. The client team brainstormed and implemented low cost ideas that ensured Winding operation will not be needed to be outsourced for a long time.

• Emerging Reality

Within a period of 100-120 days the RM, Components and FG buffers were filled. The client did not need to seek more bank loan to expand capacity. The production was planned and executed in weekly buckets making the monthly planning driven hockey stick pattern a history.

What do you think the level of revealed capacity of production at our client?

What do you think would be impact of supply reliability and speed on relationship of our client and the OEMs?

Would our client have more confidence of seeking more SKUs and entering into relationship with more OEMs?

Converting the above decisions into an Excel based Demand Driven Materials Planning Engine is another exciting experience that my colleagues Ayush Agarwal and CA Aniruddha Joshi have scripted.

?References:

1. https://www.dbrmfg.co.nz/Production.htm
2. https://www.planettogether.com/blog/why-you-should-know-the-shape-of-your-plant-vati-analysis
3. Chapter 8 TOC Hand Book - DBR, Buffer Management, and VATI Flow Classification