Goer Manufacturing Epicor Go Live

Goer Manufacturing Epicor Go Live

Goer Manufacturing was a fascinating company. They manufactured fixtures for entire stores like Sears, JC Penney, and Books a Million. This included everything that merchandise sits on, shelves, gondolas, changing rooms, cash wraps.... All of it was custom-built for each building And not just one store, but they would do out 20, 50, and even 100 stores at one time, as well as delivery, setup, and installation.

Goer was a 100 million dollar company (owned by RHC Spacemaster) with up to 250 employees over three locations, Charleston SC, Cedar City Utah, and Spartan Missouri. I served as MIS Director over all Voice, Data, ERP, and Software for all locations.

Goer was in the process of rolling our Epicor and just could not make it happen. Accounting and Management systems were up and running, but little else worked. The issues involved system, hardware, and people. Servers were crashing daily and cable infrastructure was a mess. Other than the President of of the company (who was weak on the subject), and the Chief Financial Officer, nobody wanted the new system. The project faced fierce opposition from the Sales and Production Managers. In fact, I was threatened with physical violence if I continued to push the Epicor Project.

As such, the CFO decided to put off the Epicor “GoLive” to focus on network infrastructure. In the next few weeks we rolled out new windows and hpux servers, cisco switches, new cable, WiFi and barcode. The same was accomplished for each remote location. I designed and installed the systems. I saved enough on hardware to purchase a $18k battery backup system, that supplied power to servers so that the two remote plants could continue to operate when Charleston was down. The backup paid for itself two months later when a hurricane shut down Charleston, but the backup allowed Cedar City and Spartan plants to continue production. Over the next two years, all three plants experienced near 100% uptime and availability from voice and data systems.

To prepare for the Epicor installation, I also worked 2nd shift in the mill running a panel saw. This was kind of Kool because nobody knew who I was. They thought I was just another “dust head” (their term). However, this experience provided a first-hand view of people, systems and processes in a place where the company spent the most money, where the biggest bottle-neck existed, and where the greatest gains could be made. In addition, I made some friends in the mill, right in the Production Manager's back yard, that paid big dividends later in the process.

This is what I found: 2nd shift would clock-in at 6:00pm, then take an unauthorized 10 minute break. After the break a machine operator would walk back to his machine, clean, test, and calibrate (if necessary), then walk back to the front counter to pick up one work order. The operator would use a telephone to order materials for the work order. An hour or more later, the materials would arrive and the machine would begin “making sawdust”. If an authorized break or lunch occurred, the operator would stop production, take the 10 minute break, after the break, use the bathroom, then return to the machine and resume making sawdust again. When the work order was complete, the operator would call to have the product moved, clean the machine, then take another break. After the break, the operator would return to the front desk, turn in the work order, get one more work order, order materials,... Work orders were back-flushed by the Mill Supervisor at end of day.

But wait – there's more: Daily, sales would run out to the mill and scream “Stop Production, I need a prototype!” The operator would stop what he was doing and call to have the current materials moved. Then he would take the new work order, order the materials, and go to work (if it was not break time). Fifty percent of the materials produced in the mill would be lost or damaged. It was actually in the procedures to produce two pieces for ever piece ordered to cover lost or damaged materials.

And it gets worse – changeovers were performed by the maintenance department during first shift. Sometime a changeover would take an entire day. Production would come to a halt and the operator spent his day in the break room. This sad story was repeated on all three shifts. Even so, the company managed to make lots of money.

Now that the IT infrastructure was solid, I worked with the CFO to develop a “brute force” method of implementing Epicor. The CFO and I diagrammed every process and procedure, and we trained everyone that would show up for training classes.

I spent a lot of time with the people that did the work in the departments, asking, “If you could redesign your job, this department, what would you do. If you could change how other departments interact with you, what would you change. This process took a lot of time, but we gained some valuable information and deep insight. But even more, a 'grass roots demand' was created for the new system and the benefits it would bring. That demand became so powerful, then even the Department Managers and Supervisors could not resist supporting the change to Epicor.

Several “GoLive” dates cane and went, without moving forward. Finally, the Production Manager quit. Without his strong voice, the Sales Manager switched sides and became an Epicor proponent. The next “GoLive” deadline was kept, and Epicor, finally, was a resounding success,

By “GoLive”, the opposition was non-existent and support was very high and positive. The Sales Manager and new Production Manager (the former mill Supervisor) really stepped up, learned the system, and became our biggest champions. They were determined to use Epicor to the full and make whatever changes were necessary.

Here are some of the changes made as a result of the Epicor GoLive:

Planning/Schedulers were moved to the sales department.

Sales went from 25 employees to 15

Sales could now accurately project ship dates for customers whilst placing the orders.

On-time shipments went from 40% to 92%.

Purchasing was moved to the upstairs area next to sales to promote communication.

Purchasing went form 5 employees to 2

This move meant that the purchase order were correct and materials arrived on-time.

The mill went from 3 shifts to 1.

Setups were performed during 2nd shift, resulting in no downtime for setups

Operators were given 4 hours of work orders at a time.

Operators completed work orders at the point of completion

A prototype mill was created in a former wip area to handle prototypes, emergencies, and replacement units.

Materials were moved in and out of the mill on schedule.

Processes were refined and streamlined.

Lead times were greatly reduced.

Inventory levels dropped dramatically

Miss-builds, missing and damaged product was virtually eliminated.

Production size was reduced by 30%. We rented part of the building out to another company.

Profits increased significantly.

Internal internet (intranet)

I developed what we called an “Action System”.

Anyone with a question or issue could assign responsibility and enter it into the Action.

Question and issues could be viewed by everybody with filtering capability.

The idea is that if you shine a light on a problem, it goes away quickly. If you could solve one problem a day, a week, or a month, what would that mean to you.

This system angered several Managers at first, exposing some “dirty laundry”. Bet lingering problems were solved and an amazing rate. After a year, seldom did any significant issue arrive on the system.

Production Measurements

Production and Sales found the Epicor reports very useful. But they always wondered how system numbers equated with the amount of time mill machinery was actually making sawdust.

Most of the equipment was very old (built in the 1940s) and had no way of collecting data AND we did not have Ethernet cables near the equipment. So I bought some WiFi print spoilers and modified them to turn on and off using an led light and a photo-resister. A panel saw would only make sawdust when the blade was out of the stop position. So when the led light did not hit the photo-resister, the saw was off. When the light was on the saw produced sawdust, and I could see the IP from the print server.

We collected data for 30 days, turned it into some simple bar graphs with real data in the background, and wallah, we now had instant verification of Epicor reports. It also produced ongoing data so that production and Sales could tweak the mill to maximize production.