Poster
Client Context
Cox Automotive’s subsidiary, Manheim, has two main streams of business centered around car repair: wholesale auction and retail. In the wholesale model, banks and dealerships deliver cars to the wholesale facility where Cox services them to be auctioned to third parties. In the retail model, rental companies and online used car dealerships deliver cars to the retail facility, where Cox services and details vehicles before being picked up by clients to sell or use in their fleet. Therefore, in retail, Manheim is no longer the point of sale but is an outsourced, large-scale, mechanical and body shop. The team focuses specifically on retail.
The retail process consists of 14 sub processes in five main sections: entry, inspection, parts, reconditioning, and quality assurance. Entry proceeds as follows. In Vehicle Entry, a vehicle enters the facility, its VIN number, make, model, mileage, and client are noted, then it is fitted with a GPS tracker and parked in the inbound staging area. During Vehicle Qualification, a coordinator takes the information from Vehicle Entry and answers a customer-specific questionnaire concerning mileage, warranty, and any relevant recalls. In the Inspection stage, the vehicle then goes to Mechanical Inspection in the Mechanical Shop, where technicians fill out a client-specific inspection checklist to identify any basic mechanical problems. If a technician finds a more serious problem or a check engine light is on, the car is taken to a higher level technician for Diagnosis. The car is then taken out of the shop where it undergoes Cosmetic Inspection. During Cosmetic Inspection, an inspector tallies cosmetic damage and provides labor estimates for the repair work.
Some cars require additional parts for repair and enter the Parts stage. In Parts Estimate a coordinator utilizes the scope of work from the Cosmetic and Mechanical Inspection to quote the customer for parts. In Parts Approval, the client can deny the quote, during which the Parts Estimate is repeated. If the estimate is approved, the parts are ordered (in Parts Order) after which the car awaits parts. While the car is waiting for parts, depending on its needs, it is sent to whichever sub process in Reconditioning is available first: Mechanical Repair to fix, Chip Wizard for cosmetic touch-ups and minor body work, and Body Repair for major body and paint work. After all necessary mechanical and body work is done, the car goes to Detail, where it is washed and vacuumed as the last step for Reconditioning. The car is then put through Quality Assurance. The first step, Quality Control, looks for cosmetic imperfections and ensures the completion of all work quoted and billed to the client. Manhiem then sends the summary of work and billing information to the client for approval and payment. After payment, the car awaits pick-up at the facility.
Project Objective
The complexities of the system give rise to its challenges: lengthy cycle times and increasing demand. Cox Automotive brought on Team 29 to improve cycle times, which average over 15 days. This task is made more difficult given forecasted demand growth of 20% per year for the next five years.
After collecting and analyzing data, it is clear that rework and Mechanical Inspection are the two largest drivers of cycle times. Specifically, one rework adds roughly four more days to a car’s time in the system, and two or more reworks adds roughly eleven. Because 67% of cars undergo rework at least once, Team 29’s recommendations include several process improvements to reduce the frequency of loopbacks throughout the system. In addition, Mechanical Inspection acts as a bottleneck that drives up cycle time and limits throughput. Per suggestion of Cox Automotive, the team evaluated potential changes to cycle time after consolidating mechanical and cosmetic inspection.
Despite lower cycle times and subsequently shorter queue lengths, increasing demand still challenges the parking capacity of the system. The site often experiences shortages in available parking spaces, sometimes parking cars in unmarked spaces. Unmarked spots create cascading challenges for car location and retrieval. This shortage in parking spaces, which will only be exacerbated by demand growth, creates the opportunity to design new lot layouts and increase parking capacity.
Based on these challenges, the team identified three key performance indicators: cycle time, throughput, and parking capacity. Improvements in cycle time and throughput are driven by reduced rework and process consolidation. Increases in parking capacity are developed via the design of three lot layout recommendations. The team estimated that their solutions can result in a 20% decrease in cycle time, 24% increase in throughput, and 5% to 107% increase in parking capacity.
Design Strategy
To understand the effect of rework on the system and how its reduction could drive down cycle time, the team needed to identify the areas from which rework stems, the causes of those reworks, and the amount of time those reworks incurred on a vehicle. To do this, the team analyzed vehicle entry data compiled in Manheim’s vehicle database known as RPP, which outlines the path a vehicle takes through the retail reconditioning process. From this data, the team was able to draw a representative sample of cars that received rework and could be used to apply solutions across the entire retail fleet.
After compiling the findings drawn from RPP, the team had to determine the true root causes of reworks, which could not be determined using data from RPP. True root causes are operational or behavioral happenings that would lead to a general cause. An example of this would be a technician marking down the wrong part to be ordered during mechanical inspection. Since these true root causes could not be observed via RPP, a list of hypothetical causes was drafted. This list was curated from facility observances and anecdotal data from retail personnel. Once this list was developed, a series of process improvement measures were considered. Process improvements include station consolidations, accountability and incentive measures, heightened quality assurance, and team restructurings. After high impact measures were agreed upon, the team and client reached consensus on a reduction in rework in each process by 40% to 50%. This reduction stems from controls that catch rework causes earlier and improved reporting and data to inform incentives and training.
Lastly, the reduction in cycle time these process improvement measures would generate was calculated. This was done by breaking out reworks into categories based on the amount of times they loopback through the system and finding the new average cycle time of those categories. Once determined, the weighted average of those categories and their respective proportions relative to the entire retail fleet was taken to arrive at the new average cycle time and compared against the current average cycle time.
Initially, the team’s focus, in terms of process flow, was to standardize the routing of vehicles through reconditioning processes; however, to address client requests following the interim presentation, the team shifted focus towards process consolidation. Cox noticed an opportunity for combining processes at a bottleneck of the system - Mechanical Inspection. This process consolidation involved combining the Mechanical Inspection and Cosmetic Inspection processes. Cycle time measurements and throughput were used as indicators of whether process flow could indeed be improved by combining these processes. We found that in order to realize any decrease in cycle time or increase in throughput, a 23% reduction in the additive average cycle times of the separate processes was necessary.
To meet the challenge of limited physical capacity, Team 29 designed three layout recommendations of minimum, medium, and maximum disruption and parking space capacity. After establishing space constraints for staging areas and creating basic guidelines for each layout, the iterative design process began. The dimensions and capacity of one staging area changed while holding all other staging areas constant. The team calculated the proportion of parking spaces in each area relative to one another. By comparing the correct proportions collected from work in progress data to those of the design, Team 29 measured the accuracy of relative staging area sizing. At each level of disruption, the final recommendation is the design with roughly the largest capacity and most accurate queue length proportions.
The results of the design strategy include frequency rework root causes, the effects of process consolidation, and the effectiveness of different layouts. With regard to rework, data collection resulted in a frequency table of the sub processes most impacted by rework and the overall impact on cycle time. Using this table, root causes leading to over 60% of all loopbacks were identified. Given the circumstances, the team was unable to test the potential improvements of these suggestions; Instead, Team 29 agreed upon feasible numbers with the client that comprise the value of the suggested process improvements. Using simulation software to isolate the inspection processes, the team determined that combining the processes could decrease cycle time only if the resulting inspection process time is 23% less than the sum of mechanical and cosmetic inspection process times. However, if combining these processes made the inspection more standardized and thorough and reduced rework, the team hypothesized that this consolidation could realize value through a decrease in average cycle time stemming from reduced reworks. Finally, the results of the design process show that physical capacity is limited most by the presence of storage vehicles. Once storage vehicles are removed from site, the capacity of the facility increases strikingly, even with completed staging angled parking. The potential layouts were validated by creating blueprints that comply with the system’s physical constraints, which relied on a combination of onsite measurements and online imaging data.
Deliverables
The team identified three areas of opportunity to reduce rework: cosmetic inspection, mechanical inspection, and parts ordering which over 60% of rework stems from. After a study of these processes and their contribution to loopbacks in the system, the team identified the most cost effective solutions with the highest impact.
Weekly Diagnostic Report: This recommendation addresses all potential causes of rework and gives management a high level view of what occurs on the shop floor. This report will identify the total number of cars that loopback in a week, the total added cycle time to the system, the process that triggered the loopback, and the process the car was routed to. It would also detail the cause of the loopback listed in the observation section by the technicians. This report will be used by management to devise corrective action strategies to further reduce loopbacks.
Updated protocol and inspection checklist: The team recommends an updated and more comprehensive inspection checklist based on data collected from RPP. In conjunction with updating the checklist, the team recommends standardizing the current inspection protocol. The team recommends splitting the inspection up into distinct sections based on the interior and exterior parts of the car to allow a more detailed and focused examination.
Parts & Labor estimate expert: The team recommends a parts and labor estimate expert to work on the shop floor. The expert will act as a liaison between the technicians and the parts ordering department to ensure the right parts are ordered which will avoid the aspect of guess work currently done by the parts ordering department. The team also recommends that the parts and labor expert be consulted during the cosmetic inspection process to ensure a more accurate prognosis of the labor hours required to fix a car.
Quality Control after repair stations: The team recommends that the parts and labor expert should perform a quality examination and sign off on all cars before they leave the shop floor. While this will slightly increase repair process times, it will allow technicians to fix any issue immediately. This will mitigate the overall impact on cycle time as each loopback adds a minimum of two days to the cycle time.
Continuing with the portion of the solution approach where consolidating processes could lead to lower cycle times, the team decided to deliver a recommendation for whether this consolidation should take place between the mechanical inspection and cosmetic inspection processes. A Simio simulation of the current system was completed, and then the mechanical inspection and cosmetic inspection processes were isolated. The isolation of these processes in the simulation helped to remove variability in cycle time caused by considering the rest of the system. Trying different levels of combined process time and how those changes contributed to vehicle cycle time drove the team’s recommendation for whether or not to consolidate processes. It was found that in order for process consolidation to be effective at decreasing cycle time, the combined Mechanical Inspection and Cosmetic Inspection process times would need to decrease by more than 23%. This number was found by using the simulation to calculate average cycle time given a certain combined process time. The physical deliverable to Cox was a verbal recommendation that highlights the necessary reduction in process times to realize the same cycle time and throughput as in the previously separate system, as well as a list of qualitative benefits.
By establishing the basic layout guidelines and drafting three layouts of varying disruption, the team was able to create a deliverable for the lot layout. This deliverable consists of the physical blueprints for the minimal, medium, and maximum disruption layouts. These blueprints show the boundaries of different process, staging, and storage areas at the facility, the facility borders, and a blueprint key. The key describes the meaning of each acronym on the given layout labels. The blueprints also contain information about how much of an increase in parking space capacity the given layout can provide Cox.
To add to the deliverable, the blueprints are accompanied by implementation cost and storage loss information.
Project Information
Student Team
Greyson Burnett, Charles Castle, Toffee Jegede, Joseph King, Harrison Pettigrew, Santiago Rivera
Faculty Advisor
Faculty Evaluator
