Case Study - Steel Distribution and Processing

Background

The company in this case study sells steel grades in blocks (billets) to injection molding and machine tool customers. In addition to steel sales, it also offers services such as: cutting, machining, grinding, heat treating (external service provider).

Because of these value-add services, the company is in the build-to-order business - where each quote and order is unique. This is distinctly different from a company that makes stock keeping unit (SKU) like products such as: chairs, consumer products, etc.

SKU products are not configured – in other words customers are not choosing options. This company has not yet explored nor invested in the setup of and ERP software. Like many small businesses, it is running the business using Microsoft Office – Excel and Word. Excel is used for quoting, orders, managing production orders, and other data. The company is using QuickBooks to manage the financials, material purchases, suppliers, customers.

Problem

The company has recognized that the significant manual effort in scheduling requires they do something different in how they run the business. One solution explored is an ERP software. This is a big investment and change in processes for employees. ERP software is designed to have an assembly (top level) part number for every line item (quote or order). Below this assembly part number is a unique production router and bill of materials. Using an ERP for a configured product to create a quote with 5 line items would require creation of: 5 assembly part numbers, 5 production routers and 5 bills of material (BOM). This is made somewhat easier by duplicating and then editing an existing router and BOM. The router is still a challenge however, because the process time for each of the production steps would need to be estimated and updated.

Process Steps

In this case study, the process steps are typically:

---

Heat treat and grind are optional, depending on customer requirements. The bill of materials presents another problem.

BOM

The cost of the material is a function of:

1. Grade
2. Volume (cubic inches). The volume is a function of:
   1. Length
   2. Width
   3. Height

Out of the box the ERP would require each unique combination of these four features – grade, length, width, height. One could allow a tolerance on the L, W, H to minimize the number of part numbers required to cover all possible combinations. Still, if each is 5-50 inches in steps of 1 inch – the combinations if there are 5 grades is: 5 x (50-5) * (50-5) * (50-5) = 455,625! A workaround might be to use volume rather than dimensions. The number of part numbers required would be 97336 (table, right). Further reduction of the complexity of materials would require customization ($$$) in the ERP software.

---

An alternative would be to manage the inventory (and cost) separately (Excel or other software), and enter the material cost manually. While less complicated, it introduces the possibility for error which can directly affect order profitability and profitability of the company.

Router

Switching focus to the router, the number of combinations would be:

1. Cutting – NO, YES (2)
2. Machining – NO, YES (2)
3. Heat Treat – NO, YES (2)
4. Grinding, NO, FinishQuality1, FQ2, FQ3 (4)
5. QC/QC – YES (1)
6. Shipping – YES (1)

...resulting in 2 x 2 x 2 x 4 x 1 x 1 = 32 combinations – a manageable number. Some bad news - we have not yet considered the fact that each of the steps that are not “NO” would have different process times (minutes) that are dependent on the material chosen. You can imagine that the time required for 1) Cutting, 2) Machining, 3) Heat Treat, 4) Grinding would vary based on whether the hardness of the steel. The QC/QC step would depend on grade and weight (volume * density) and Shipping would depend on weight. Because the times for a router step are unique to that router, this means that 32 is not the final number. The final number is 5,000,000 (five million!):
Grade (5) * Cutting (10) * Machining (10) * Grinding (10 * 10 * 10 * 10) * QA/QC (1) * Shipping (1)
Clearly the feasibility ends at manual creation of a few thousand routers (fewer is better!). Anything above that is not feasible. 

---

Solution

Velosity solves this problem uniquely. Velosity does not require the creation of: 1) assembly part numbers, 2) routers, 3) bill(s) of material. Instead, Velosity generates the router and BOM dynamically based on the product features (attributes) selected. To simplify this example, the MACHINING step is left out. MACHINING would require definition of machining types: 1) volume removal for facing, 2) volume removal from boring of various diameters, etc.

The "magic" of Velosity is defined in four parts:

1. Attributes (Product Specific)
2. Items
3. Tasks (Process Steps)
4. Rules (Product Specific)

1) Attributes

The attributes that the user would select are shown here. These are setup in Velosity.
Shown below are the Attributes as setup in Velosity

2) Items

The items that are added by the Rules or Item Mapping are shown here.
Shown below are Items as setup in Velosity

3) Tasks

The process steps (tasks) for this company are:

• CUTTING
• HEATTREAT (External process)
• GRINDING
• RECEIVEINSPECT
• GRIND

Shown below are the Tasks as setup in Velosity

4) Rules

The rules that are evaluated to generate the router and BOM are shown here.
Shown below are Rules as setup in Velosity

Results

With this setup, the customer is able to generate quotes and orders within a few minutes (without creating any routers or BOM!), like this:

Conclusion

Velosity transformed the quote and order management processes for this company compared to Excel being used previously. It also gained the ability to quickly lookup historical quotes and orders.

The company is next looking to use the production order and scheduling functionality in Velosity to reduce the effort required for scheduling, completion date visibility. Part of this improvement will be the rollout of rugged tablets at each workstation to provide shop personnel with visibility of current and upcoming production orders along with production order time clocking. The time clocking will provide data to improve the time calculation in the router steps – leading to new insights about efficiency and pricing vs value delivered to their customers.