Spring 2023 Senior Design Expo
Spring 2023 Projects
Eaton is a power management company that produces electronic breakers for worldwide clients. Their warehouse in Beaver, PA contains over $26,000,000 of inventory and storage was overcrowded, unorganized, and lacked discipline. This caused slow and inaccurate “kits” to be sent to production lines, which slowed production as a whole. Eaton contracted our group out to solve this issue. Through Eaton’s and our own collected data, we determined a set of root causes through a fishbone analysis. After this, baseline rates for picking inventory and inventory accuracy were calculated. We then used principles of industrial engineering to develop a set of recommended solutions to target these root causes which will be soon adopted by Eaton’s production plant. These solutions included a full new layout with permanent locations for over 1,400 different parts, a new area designed specifically for creating kits, and new carts designed with ergonomics in mind for the employees who pick the inventory. This permanent layout used our group’s ABC analysis as well as principles of material handling equipment to create an intuitive, easy to learn storage solution. With new roles came a new process for picking, so our group then created a new process flow map. With these new recommendations, we were able to cut wasted movement down by at least 60%, resulting in a greater pick rate. With the new layout and a proposed control plan, the inventory accuracy would increase by over 40%, resulting in accurate kits arriving at the production lines. After creating our deliverables, we conducted a cost analysis for the project and determined a return of investment of 2.6 times the investment after five years.
Eaton corporation is power management company located in Beaver, PA. Currently, their NRX production line, a line producing two types of low voltage circuit breakers, is inefficient, utilizes poor material management, has unnecessary operator travel, and has a low production output. Our team worked closely with Eaton’s operations team to identify the causes of these inefficiencies and problems.
The current line is configured in a U-shape with operators positioned on the outside. By first moving them to the inside of the line, workers will be able to flex and switch to different stations efficiently. The breaker accessories and kitting areas can be moved closer to the main line to reduce the amount of time spent walking by operators. Additionally, machinery that is favored by operators can be moved into the main production line to reduce breaker travel time originally done via rolled carts. There is one main bottleneck station early in the process that leads to high work in process between stations. The Senior Design Team also concluded that it would be in the best decision to purchase additional machinery at this test station to increase throughput to other stations on the line. The additional machinery would add an additional operator to operate the line, but this would produce a return on investment of only 8 months due to the profit from increased production of circuit breakers and reduced travel time.
FedEx Ground is currently experiencing long lead times for spare parts that has led to increased network risk, downtime, and lost savings as the procurement team does not have the necessary time to secure the best prices. The current process for acquiring the spare parts goes through five different departmental teams before ordering can begin. If anything is missing from the list along the way, it is sent back to the supplier and the entire process starts over. This can take several weeks to receive the completed list and takes away the procurement teams' time searching the market for the best prices, forcing them to purchase all the parts from the integrator. Our approach to the problem was to focus on what can be fixed on the FedEx side of the process, as we have no control over the integrators or suppliers' actions. Our suggestion is to introduce a two-list process that will allow the procurement time to have more time to look for more parts and reduce the number of parts they must purchase from the integrator. Using a statistical analysis, a classification model, and a simulation model, we can model the current process and compare it to our proposed solution using cost savings and max lead time as our metrics. Using these metrics, we were able to determine an increase in cost savings of $1049.95 per list and a decrease in max lead time of 2.67 weeks by implementing our prosed solution. Our final recommendations are to implement this two-list process as the standard workflow and utilize the classification model to predict future lead times for both short and long lead time parts.
FedEx Ground is the low-cost ground shipping service that prides itself on being a faster alternative to UPS Ground with shipping times within the United States. With facilities nationwide, FedEx Ground relies on conveyor systems to handle and sort packages in each facility. FedEx Ground has unplanned downtime events with their sorting facilities that delay order fulfillment times and cost the company millions of dollars. Additionally, FedEx Ground has recognized that their current method of tracking such issues is insufficient. To address these issues, the team successfully created two new metrics for the company to implement immediately into their analysis. The new metrics allow for a clear comparison across different facility sizes, and improve the understanding of a downtime event. Additionally, the team recommends the implementation of Domo, a reporting software to improve communications and analysis of events. Finally, the team found that equipment damage and failure was responsible for 34% of all downtime events experienced. Therefore, the team recommends the Waites monitoring system hardware in order to monitor the conveyor systems and ensure the conveyors are maintained to prevent failure or damage during a sort. The hardware has the capability to measure temperatures and vibrations experienced by the equipment to alert workers if the equipment undergoes any abnormal conditions that may cause damage and failure to the system. With proper implementation of the hardware, the company has the potential to see approximately $2 million of cost savings per year.
FedEx Supply Chain is a third-party logistics provider that will soon be taking over the Consumer Cellular warehouse where phones are activated with SIM cards, connected to the network, and bundled with information sheets. Orders to retail distributors are large with short fulfillment windows and as a result, they currently operate a “push” system in which work is completed before receiving orders. This leads to a large build-up of finished inventory. Further, the steps to activate and bundle the phones are not as efficient as they could be. To achieve FedEx Supply Chain’s goals of less labor costs and lower finished inventory levels, the team’s work is split into two segments: work process improvements and scheduling policy optimization. For improving work processes, the time-motion study methods of Therbligs and MOST were conducted on a video received from the client. These analyses resulted in the generation of six time-saving recommendations. All together, these recommendations improve the productivity rate of phone activation and bundling by 7.5% to 14.4% resulting in an estimated savings of about $97k over the next 3 years. Standard Operating Procedures were made to communicate these recommendations. For scheduling policy optimization, demand was simulated and various policies tested. These policies decide how much to produce in a day, which phones to produce, and which orders to satisfy. They include parameters such as number of workers, number of workstations, limits to overtime, and the number of cross-trained workers. In total, over 150k years of fulfilling orders were simulated. The greatest impact found was that having flexible cross-trained workers is critical to satisfying orders on-time with low total labor costs. Recommendations were also made about other parameters such as the amount of buffer needed and which phones can be prepared before receiving orders for them. Overall, our work saves on labor costs and will minimize finished inventory levels.
The goal of this project was to develop an optimal and feasible layout for General Electric’s new warehouse in Charleroi, PA. The team’s first step in the project was to visit the current warehouse site to better conceptualize and understand the inventory processes that take place throughout the warehouse.
Currently, there is a high volume of material coming into the current inventory warehouse, space is not optimized to handle increasing demand, and the current warehouse is not optimized for a first-in-first-out (FIFO) system. These issues lead to productivity issues, kitting errors, and difficulty supporting increasing demand.
Before alternative layouts could be developed, it was necessary to evaluate overstock, material volume, department locations and relationships, and the current warehouse capacity level.
Once these facets of the project were analyzed, alternative layouts for a new and larger warehouse were developed. These layouts include adequate shelving to support increasing demand. They also take into account the relationships between departments (such as the receiving bay and floor storage area).
The two alternative layouts were evaluated via rectilinear distances between departments and layout efficiency ratings. Ultimately, it was determined that Alternative 2 would best serve the needs of GE.
In addition to the development of a warehouse layout, visual management and material handling solutions were identified. Some of these solutions include a barcode-by-bay system, large overhead signs, the application of RF Scanners, and the acquisition of 1 Clark scissor lift and 4 Clark narrow aisle forklifts.
IDL is a global retail design and project management company that is part of Matthews International Corporation. It provides design, engineering, fabrication, fulfillment, and installation services to retailers and brands. Its services include retail strategy support and design, consumer and brand insight, retail re-imaging, concept shops and pop-up retail, custom store fixtures, permanent and temporary displays, program graphics, product launch support, and brand development and execution solutions. Our task was to help IDL improve its packaging labor source. Currently, the packing line utilizes operators to package signs without the use of automation. In peak seasons IDL currently does not have enough workers to keep up with demands and has to outsource temporary workers. This would lead to delays and efficiency due to inexperience on the line and packing process. It would also lead to more expenses for IDL. To help solve the IDL problem our team researched and reached out to vendors to see what types of automation can be used to help the manual packing line. This would lead to our biggest constraint as communicating with vendors can take days or even weeks. After we reached out and got information about the types of machines, for example, MKW collators, Technopacks shrink wraps, or Pineberry friction feeders we calculated the ROI and incremental analysis. IDL also wanted us to find the ROI for these machines and the incremental analysis was another source to help justify which machines to get. From our results of the ROI, the technopacks shrinkwrap had the best return on investment and for the incremental analysis, the MKW had the best present worth value. While IDL only wants to select one of these machines if possible we would recommend both of these to them.
IDL Worldwide is an international printing company that services the needs of large commercial brands such as Chick-Fil-A, Sheetz, and Giant Eagle. IDL currently uses large and small format printers in their operation, as well as manual die cutters and digital cutters or plotters. The problem that IDL is facing is a lack of information and analyses required to make informed decisions regarding the purchase of new printing and cutting equipment. To solve this problem, the team completed return on investment analyses for new printing and cutting equipment, facility layout analyses on the proposed printing area and new locations for a plotting machine, and constructed a simulation model, to perform a sensitivity analysis, on the new printing equipment.
Upon completion of the printer and plotter ROI analysis, it was determined that the Konica Minolta (KM) printer yielded the highest return on investment with an IRR of 57% and the Elitron plotter yielded an IRR of 42%. The group then constructed new layouts with the KM printer, increasing adjacency ratings from 45% to 78%. New plotter locations were also analyzed, and travel distance was ultimately reduced by 28%. Finally, upon completing the sensitivity analysis on the KM printer using Simio modeling software, it was determined that, even with a 50% increase in demand, the KM printer would only reach a utilization rate of 44%.
Our final recommendations to IDL included purchasing the Konica Minolta printer and Elitron plotter. In addition to buying these machines, IDL should adopt new layouts that increase adjacency rating and decrease travel distance. IDL will be able to meet demand and will have ample room for growth in a market that is expected to grow by 6% in the next year.
JADCO is a family-owned manufacturer of supreme quality impact and abrasion resistant steel products. Their manufacturing facility is located in Harmony, PA and was the main point of focus for this project. The current layout of the Upper Weld Shop of the Chromeweld Department was designed without prioritizing machine relationships. This created long path lengths and excessive non-value added time in the steel production process. Our project goal was to produce two alternative layouts that have a direct process and material flowpath. This in turn would minimize backtracking during production, decrease travel distance, and reduce costs.
Our two alternative layouts consisted of an optimal and feasible layout, and both were visualized using AutoCAD. The optimal layout was determined using systematic layout planning and the DMAIC approach to problem solving. The feasible layout was determined based on client feedback, and prioritized minimizing machine movement. Both layouts had high efficiency ratings as well as a substantial decrease in path lengths. Furthermore, an in-depth cost analysis of each layout was performed and presented to JADCO. Labor savings, potential increases in revenue, and simple payback periods were calculated in order to justify implementing our suggested layouts. In addition, the potential increase in production in terms of parts per day was calculated for each layout. We concluded that our suggested layouts were favorable based on the results of this analysis.
Regarding a long term implementation plan, we recommended that JADCO look to incorporate small changes (e.g. adding additional storage), and consider adjusting the layout as machines naturally need to be replaced. We also recommended that JADCO consult a construction company for better estimates of the time/cost of implementation.
MSA develops, manufactures, and supplies safety products that protect people and facility infrastructure internationally. Within recent years there has been limited visibility and incorrect categorization of MRO spending. MRO stands for maintenance, repair, and operations thus, items defined as such are not a part of the finished product but are used during the production process. Systematically at MSA, an employee purchases a product and then links said purchase to a G/L Account to categorize their purchase. Employees end up linking purchases to the wrong account if they input the incorrect number, which is a common occurrence. The result is incorrect spend data leading to difficulties in leveraging vendors, obtaining spending overviews, and budgeting. To mitigate these issues, we came up with a two-part solution plan. The first part looks to improve the spend data collection process through a vendor selection catalog system on Microsoft Teams, the platform MSA uses company-wide. The system is comprised of past purchases from MSA’s preferred vendors such as Amazon, Fastenal, and Uline. The purchases are sorted into G/L Accounts folders from three broad categories Repair & Maintenance, Supplies & Materials, and Other Functions, and put on spreadsheets. Through this method, MSA can standardize their MRO purchase process. The second part was creating a Tableau dashboard that displayed MRO spend by plant G/L Account, and overall monthly spend each with indicated budgets. We recommend that MSA use a monthly MRO budget that is five percent of their total procurement budget. The company should also continuously update the spreadsheets of purchase with new products and vendors, as well as contact managers of plants that are superseding the budgets of other plants.
UPMC Shadyside is experiencing longer than expected inpatient hospital stays that affect the profitability of the hospital, the quality of care for patients, and the number of patients that they can treat. Centers for Medicare & Medicaid Services (CMS) calculates the geometric mean length of stay (GMLOS) for each Diagnostic-Related Group (DRG) designation. When patients overstay this CMS standard, the hospital is not only unable to bill payers but also cannot begin treating a new patient due to limited bed capacity. In 2022, 57% of patients at UPMC Shadyside stayed longer than the expected GMLOS for their specific diagnosis. Our project goal was to locate areas causing the most substantial impact to inpatient discharge delay, uncover reasons behind the process bottlenecks, give recommendations, and configure methods for continuous improvement.
In order to address this problem, the project team interviewed UPMC staff familiar with the process and analyzed historical patient data to better understand the factors which affect the length of stay. One such factor, we hypothesized, was consulting physician turnaround times.
Mainly the effect of consults on a patient's length of stay was examined. The team examined patient data for cases with and without consults and grouped the cases by DRG to eliminate case complexity as a possible confounding variable. The analysis showed a difference of almost 1.5 patient days in some DRG categories and indicated there is likely a relationship worth further investigation.
Additionally, the team developed a machine learning prediction model using data available early in a patient’s stay to better determine when a patient’s discharge should be planned. This was intended as a proof of concept to show how new techniques, including machine learning, could be applied to the healthcare industry.
Based on qualitative observations by members of the UPMC staff embedded in the process, the team made recommendations to standardize practices in laboratory/radiology orders and elevated care facility placement. By implementing our recommendations, we estimate a reduction in patients’ length of stay, an increase in the volume of patients seen, and greater overall hospital profitability.