Savvy students help shipping department move …
|FROM THIS ...
||TO THIS ...
|FROM THIS ...
||TO THIS ...
|FROM THIS ...
||TO THIS ...
By Shahrukh A. Irani
The organizational hierarchy of any company has executives in the C-suite, vice presidents, managers, engineers, supervisors and employees. Yet much of the literature suggests that continuous improvement events (kaizens) primarily are done by shop floor operators and office staff. Typically, maybe the supervisor and employees who work in the department, line or work cell will be involved in that functional area’s kaizen. Otherwise, the staff won’t accept improvements and the ideas will be implemented, at best, half-heartedly.
But shop floor operators and office employees often grapple with technically demanding problems that necessitate the use of IT, data mining and other engineering expertise. So we should expand our membership choices for problem-solving teams so that work experience combined with computer skills and analytical abilities help to solve a complex problem. “Together we achieve more” ought to be the guiding rule for team formation to solve a pressing problem, regardless of how hard or complex that problem is.
Additionally, many practitioners fail to use a resource that is available, eager to help and well-educated: industrial engineering students. My bias toward these students was formed during 22 years in academia before I started working with Hoerbiger Corporation of America (HCA-TX). Industrial engineering students are a vital missing ingredient that should be added to different continuous improvement teams working with professional industrial engineers.
Into the real world
The ability to test this claim in the real world came about after I met with HCA-TX shipping supervisor Charlotte Pett. She welcomed the opportunity to add a student IE to her team. Our guidelines for forming problem-solving teams came to include:
- Planning the composition of the team to suit the scope and complexity of the problem
- Prioritizing the problems based on impact to the bottom line, not who will be available to work on the team
- Selecting measurable key performance indicators that directly relate to safety, quality, on-time delivery and cost reduction through waste elimination
- Involving external resources whenever possible, which can include student interns from a reputable industrial engineering department
- Leveraging the expertise of on-staff industrial and manufacturing engineers
- Utilizing computer-aided analytics, such as Six Sigma statistical analysis software, if the problem merits it
Thus, the core of “team shipping” became a well-rounded group that included Pett, employees, maintenance, a warehouse and inventory supervisor, a co-team leader, a materials manager, an industrial engineer and a facilities and maintenance manager. This department is the closest to the customer, and its main objective is to maximize the dollar value of shipped orders every month. According to Pett, cramped and chaotic conditions contributed to most of the inefficiencies that affect the monthly dollar value of shipments.
Cramped workplaces suggest that space is being wasted. Thomas Leskowschek, an undergraduate intern from Austria who is studying for his B.S. in industrial management at the FH-Joanneum University of Applied Sciences, joined the team as we decided to assess how much of the current floor space in the shipping department was “dead” and could be reclaimed. The IEs led team shipping on a gemba walk through the department and pointed out examples of areas that were value-added, necessary but not value-added, and non-value-added.
Since the non-value-added “dead” areas were occupied by junk, the first continuous improvement (CI) project was two hours of simple housekeeping. Everybody stuck red stickers on items they were confident nobody had any use for. The pile of junk eliminated from the shipping department played an important role in convincing them of the merit of a lean journey.
This sort phase of a full-fledged 5S program usually yields quick visible results. But the real benefits come from its elimination of hidden evils, such as ergonomic risks, inventory costs, and inefficient flows of people and material.
After the “first S of the 5S” (sorting) event was complete, team shipping did another walk-through that pointed out where floor space was wasted in the current layout. An area along one wall was “locked out” by three shelves carrying dead inventory, unused bins and incomplete orders that had not been shipped. Unknown to us, there was a rack outside the shipping department that also contained dead inventory of corrugated materials used to make shipping cartons.
As practitioners know, for CI/lean to succeed, management must empower and incentivize the workforce. Shipping employee and team member Juan Nunez came in on a Saturday for an impressive solo CI project. He emptied the outside rack by moving some of the dead inventory to the top shelf or relocating it to the “barn.” He did the same with the pallet on the floor next to the rack.
Next, he categorized all the incomplete orders on the inside shelves that were waiting to be shipped (i.e., kits for General Electric, rings, packings, quick response cell). Each category of shipments was placed on a particular shelf on the outside rack. The QRC parts, which are shipped in small white boxes, were moved to a separate multishelf cart on wheels placed next to the rack. Next, he printed paper labels for each category and taped them to the appropriate shelf on the outside rack.
Later, on the advice of co-team leader Armando Gomez, who completed a 5S project in the power rings cell, he replaced these paper labels with magnetic labels made in-house.
Subsequently, Nunez eliminated a large table inside the department by finding a new home for it elsewhere, where it was put to good use. Finally, along the opposite wall, which had many shelves full of bins that store a large variety of packing rings, he did another sorting effort that consolidated inventory. This emptied an entire rack and its bins.
Leskowschek, who was conducting time studies in the power rings cell, advised Nunez to send this rack and its bins to that cell. There they are being used for machine-side storage of scrapped metallic rings. These rings, when placed on the magnetic table of either Blanchard grinder, help enclose nonmetallic rings that would otherwise fly off the machine table during the grinding cycle.
The shipping department at HCA-TX is essentially a job shop. Each of the five cells in the machine shop is high-mix and low-volume. Both molding cells (cold compression molding and hot compression molding) are high-mix, low-volume flow shops. These cells make items that they send to the shipping department, which packages and sends them to customers in many countries, which results in different workflows. Also, the high mix of kits has different packaging requirements due to carton size, container type (corrugated or wooden crate), labeling specific to the customer and country of destination.
With her experience and knowledge of the diverse workflows, Pett listed the following products with workflows that dominated the material, person and information flows in the shipping department: packing rings, piston and rider rings, QRC packing rings, bushings and cases, and kits for General Electric.
Leskowschek used the information provided by Pett to develop the PFAST input file, which is a spreadsheet submitted to the PFAST (Production Flow Analysis and Simplification Toolkit) software that helps implement lean in job shops. Pranav Joshi, a graduate student from the Department of Integrated Systems Engineering at The Ohio State University, processed the data and emailed the PFAST analysis report to us.
The team used these outputs to generate five new alternate layouts for the shipping department. These layouts were designed based on different criteria, such as a separate cell for each customer, a central shared IT hub and other desired features listed by team shipping. Figure 1 shows the current layout, while Figure 2 shows layout No. 3, the one that team shipping selected as the starting point for the new department layout.
After Leskowschek left, Texas A&M industrial and systems engineering graduate student Clement Peng helped team shipping on this project. His independent study project during the spring 2013 semester aimed to develop a detailed blueprint for the final layout, including a budget and implementation timeline. He visited HCA-TX every Friday during the spring semester.
In the real-world classroom of industry, lean (aka “Toyota IE”) radically changes the standard approach taught in contemporary facilities planning textbooks. The footprint of each workstation, table, aisle, rack and container in the layout is potentially bloated with waste.
Toyota either pioneered or raised the importance of concepts such as right-sizing, mobile machines, reconfigurable layouts, visual WIP management, combined operations, jidoka (automation with a human touch), parallel operations, batch streaming (one-piece flow or transfer batches) and more.
For example, on the packaging table, shipping employee Robert Lu places all the parts that are going to be shrink-wrapped on a skin board (cardboard backing). Next, he picks up the skin board, slowly turns around and places the kit on the table of the shrink-wrap machine. Should not the two tabletops that he works on be a single sliding table that slides into and out of the shrink-wrap machine? At least that idea made us all pause and think for a moment during one of our weekly team meetings.
Since inventory costs are visual and measurable, team shipping decided to take a systems approach to control the purchasing costs for its inventory of cartons, which are used to ship products worldwide.
Examining the area revealed that HCA-TX had excessive inventory for a number of carton SKUs. Pett and Leskowschek collected data on purchases of the different SKUs made from June 20, 2012, to Nov. 7, 2012. This time series plot of the data did not yield any insights. But plotting the same data using the classical Pareto rule of 80-20 (see Figure 3) yielded valuable insights.
Now here is where what meets the eye might not match reality. Pett pointed out that the high inventory of GE skin boards was unavoidable because we used that item to package shipments to sister divisions and the supplier would ship only full pallets, not partials. So what appeared to be inventory waste was necessary and unavoidable. Still, these discussions led to the decision that, since replenishments were done every week, the purchase quantity of every SKU ought to be equal to the demand forecast for that week plus buffer stock.
In addition, a team of employees came up with a “bicycle rack” design for storing the unused inventory of cartons, installing it on a Saturday. This design exploits the natural shape and size of the cartons, compacts all of the inventory into a smaller volume, and makes it easy to eyeball each slot to know how much on-hand inventory we have.
Another idea being pursued is to develop a demand forecasting model that uses the past few weeks’ consumption for any SKU to estimate the requirement for next week. Of course, Pett, who has intimate knowledge of the shipment schedule for the next week, will adjust this forecast. Shipping employee Darrion Wilson is collecting the weekly usage and receipts data for each SKU and IE Shalini Gonnabathula is tweaking the forecasting model using Excel. While many practitioners maintain that there is no role for computer-aided analytics and IT in lean implementation, sensible and pragmatic integration of theory and practice yields far better results than using just the lean tools, gut feel and practical experience.
The data being collected will help sort the SKUs by volume into runners (high weekly usage) and strangers (low weekly usage). This, in turn, will help implement a one-bin kanban system for any SKU that is a slow mover. The entire inventory for that SKU will be stored in the outside rack. A two-bin kanban system will be used for the runners.
Each of those SKUs will be split between a tabletop storage rack with two days usage and an outside rack that will hold three days usage. Alternately, a mobile carton stand will be kept next to the table where Wilson packs cartons.
The development of this visual inventory control system received valuable input from a supplier representative. One day I saw the representative using his iPhone to swipe bar codes for items stored in lunchroom cabinets. That was my first introduction to a vendor managed inventory system that starts as a bar code swipe and ends as an order quantity for that item, which is entered into his company’s ERP system.
We discussed our mutual interest in lean, and he showed me an e-kanban system used to manage supplies in the first aid cabinets on the shop floor. He placed numbers on the red bins that connected to the bar code for that item on the sheet stuck inside the glass door of the cabinet. Soon after, we started building a system to automate the weekly replenishment of three of the cardboard carton SKUs.
Once that system is debugged and fully operational in the shipping department, all that will remain is to implement the same data-driven computerized inventory control system for QRC packing boxes, GE skin boards, wooden cartons, packing rings, powders and bar stock. Who said lean is hard?
Now for the hard part
The downside of having quickly plucked many of the low-hanging fruits so soon on the shipping department’s lean journey is that the complex problems now need to be tackled.
Figure 4 presents the potential for reducing the number of different carton sizes that HCA-TX buys. Standardizing and reducing the number of sizes used could help reduce purchasing and inventory carrying costs. For example, carton sizes 8-by-6-by-6 and 6-by-6-by-6 differ by only 72 cubic inches.
Let’s study the data on the usage of these two sizes during the period from Oct. 9 to Nov. 6. If HCA-TX used only the larger size, the company would ship a total empty volume of 2-by-6-by-6-by-56 cubic inches filled up with crumpled paper or foam padding. How does the cost of doing that trade off against buying 35 more cartons of the 8-by-6-by-6 carton size and not holding any safety stock for the 6-by-6-by-6 carton? Such a project that combines technical and economic issues could be offered to an IE graduate student for a master’s thesis.
Meanwhile, Figure 5 shows how the on-hand inventory of the many different packing rings made and sold by HCA is distributed in bins kept in floor-mounted racks (Q-bins) and the Space Saver, which is an older generation vertical lift module or automated storage retrieval system. We think that the current packing efficiency of these two storage areas could be improved, especially in the case of 38 kits where each kit contains a set of different rings. In each of these kits, the heavy metallic rings in the kit would be stored in the Q-bins, whereas the lighter nonmetallic rings in that same kit would be stored in the Space Saver. This project has two interlinked goals: First, reduce the average pick time to collect the different part numbers that constitute a multicomponent kit, and, second, reduce the on-hand inventory levels of these rings. Currently, this is being undertaken by Joshi, the OSU graduate student, in partnership with HCA-TX warehouse and inventory supervisor Andrew Reynolds.
Finally, remember the rack outside the shipping department? The department is assessed on just one KPI – total dollar amount shipped per month. Going by that metric, all the dead inventory and orders on that rack should not be there. Discussions with a cross-functional team composed of Reynolds, materials manager Anthony Herrell, Pett, Shane Baldon from customer service and machine shop supervisor Greg Oakley helped identify several reasons orders end up on that rack.
This is a project that could interest students looking for a green belt Six Sigma project. Other options include using green belts on staff at Hoerbiger’s Pompano Beach, Fla., location who could partner with the HCA-TX resident IE. It remains to be determined if there is enough bang for the buck for this project.
Balancing theory and practice
Like many small manufacturers, Hoerbiger is not geared to solve complex operational problems as if it was a three-year research project funded by the National Science Foundation. The shipping department is a high-pressure work environment where employees may be time-constrained and labor-constrained, but they do not lack the patience and determination essential for their job. Everybody in that department has one daily goal – receive the stuff coming in from one door and get it out the other door onto a truck that same day to make our customers happy. If operational problems arise, they solve them using common sense, firefighting, rules of thumb, resignation, brute force, overtime, teamwork, negotiations and prayers, too.
But there is considerable benefit for such companies to establish university-industry partnerships with several industrial engineering departments to enhance their problem-solving and continuous improvement project execution. Toyota is living proof that lean is the correct industrial engineering that should be taught to current and future IEs.
Much of the contemporary industrial engineering being taught in curricula across the world is irrelevant or unsuitable for commercial/real-world implementation. More than anything else, we need more industrial engineering faculty to enter industry to gain work experience and conduct applied research to advance industrial engineering practices. HCA-TX is striving to create a learning environment that teams its employees with industrial engineering students and faculty who wish to learn the concepts, methods and software tools to implement “lean IE” in every aspect of their production systems.
Shahrukh Irani is the director of industrial engineering research at Hoerbiger Corporation of America. Previously, he was an associate professor in the Department of Integrated Systems Engineering at The Ohio State University, where he received the Outstanding Faculty Award for excellence in teaching from the graduating classes of 2002-2006 and 2009. In 2002, he received the Charles E. MacQuigg Student Award for Outstanding Teaching from the College of Engineering. His research at OSU focused on developing new industrial engineering methods to adapt and scale lean for use by high-mix, low-volume small and medium enterprises.