The client is a multinational conglomerate that focuses on industrial engineering and steel production. PMI offered a discrete-event simulation model to evaluate the design of a new production line and validate the throughput capacity envisioned by the client. Our utilization of discrete-event simulation techniques allowed the client to test different layout and process configurations in their design phase of their project.
The facility has four operations – two assembling operations, one press area and one assembling/testing process split into various work stations.
One operator replenishes raw components for the first two operations and there is one robotic arm in each of the operations to transfer parts between the work stations.
The press area contains one press that receives assembled parts from previous operations and compress them so one of the three operators, on the last two operations, can pick and transfer the compressed part to the last operation.
The fourth operation has one welder, one conveyor, three work stations and three operators who will finish assembling the parts and perform various tests accordingly.
The current line design was not finalized and had not been tested to see if it was able to meet customer demands in terms of volume, cost, and quality. Hence, there was a need to simulate the different operations to identify any design problems, equipment utilization, headcount and overall throughput capacity.
The data and layout provided by the client was imported to SIMUL8®. The four different operations were included in the model. An Excel® interface was created to input data for the simulation model. This unique technique by PMI allowed the client to have the flexibility of changing most of the inputs for the simulation directly from the Excel® interface, reducing the modeling.
The discrete-event simulation model successfully and accurately determined the overall throughput capacity of the given production line design as well as the utilization of the different operators and equipment. Using the results from the baseline model, process improvements were made to the original production line. These improvements were then tested by running the simulation model for multiple scenarios. The results were used to find the best configuration that would maximize the overall throughput capacity and reduce the headcount.
In addition to the simulation study, an Excel® interface was provided to the client for making changes to the operation times, which will allow them to run what-if scenarios in case the process specifications change. Additionally, by using the simulation model to test different layout and process configurations, the client reduced the headcount by one and the number of tools used on the last operation by two. Furthermore, the client also found the best way to use its resources and maximize the line production capacity. The ROI on this project was 10 times the amount invested on the simulation study.
PMI was retained by a major automotive OEM to perform analyses of material flow within an integrated stamping and sub-assembly plant. The OEM not only wanted recommendations on proposed bar-coding systems and reallocation of production personnel, they also required a reliable tool with which to evaluate future proposed changes to the system. Throughout the project, PMI’s team utilized a variety of industrial engineering techniques. Recommendations were offered, and a custom fit tool was created. Through use of these instruments, the client’s requirements were met.
Inefficient material flow
Inefficiencies in storage areas and storage requirements
Outdated databases and standards
Insufficient reporting system for maintenance scheduling and bar code scanning
The plant studied was one of the largest in the automotive industry, containing 23 press lines and occupying 2.5 million square feet. Key system details included:
Stamping lines’ output passed through the sub-assembly area before being shipped out of the plant
Material flow was generally ‘linear’ – entrances and exits occurring at opposite sides of the plant
Stored materials were housed in containers or racks
Forklifts and dolly trains were the main form of material transport
The plant was suffering in several areas relating to inefficient material flow:
More forklift operators than necessary
Inadequate storage areas
Ineffective bar code system
Inadequate system for reporting equipment utilization and maintenance scheduling
PMI’s plan was to thoroughly study, analyze, and evaluate infrastructure requirements for better tracking and management of the material handling equipment fleet in the plant facility. This was achieved by utilizing several methods including: continuous and elemental time studies, static simulation modeling using Flow Path Calculator; and dynamic simulation modeling using Witness software.
Upon project completion, PMI’s team delivered:
Headcount reallocations: The plans exceeded the initial goal of 22 operators reallocated
Simulation models: The analytical tool allowed for quick analysis of material handling resources required by changing production conditions in the plant from both short-term and long-term changes to the production schedule
Bar Code and ID System Analysis: Full alternative, decoupled solutions that could be pursued in sequence or in parallel
PMI’s solution offered tremendous savings to the automotive OEM:
Headcount reductions resulted in an annual savings of $4.3 Million.
Bar Code and ID systems recommendations totaled $1.3 Million in potential savings.
A large discount retailer was preparing to incorporate a demand-driven scheduling system. A key parameter required for this system was accurate workload content by task for each individual department. This is a classic Industrial Engineering function and the retailer employed PMI to propose a methodology and to execute the study. While collecting this data, it was important to use Lean principles to identify opportunities to reduce waste and suggest process improvements.
Six stores across two states were studied. Within each store, three departments were studied. The departments studied were Lawn and Garden, Stationery, and Toys. The toy department consisted of the retail floor as well as an assembly area for bicycles. There was one common set of tasks which was applicable to all retail departments and a separate list of tasks for the assembly area.
The demand-driven scheduling system is highly desirable for the retail industry because it is crucial to provide customers with the desired service level, while avoiding overstaffing. Lean principles are currently finding their way into industries outside of manufacturing and the retail industry is no different. By identifying waste within a store, processes can be streamlined and process times can be minimized; thus improving the customer’s shopping experience and minimizing the associated costs to the retailer.
PMI utilized random sampling to measure the workload within each store. The study encompassed one business cycle across six different stores. A business cycle was defined as a seven day period, all hours of operation, as well as the opening and closing activities of associates.
Random sampling data was used to develop standard times for tasks. The number of items sold was used as the workload driver for each department. This data was used to develop demand-driven schedules. Several additional analyses were performed using this data including:
Analyzing the impact of government regulations for applying price tags to items versus shelves
Investigating the results of scheduling department managers during peak hours
Comparing task proportions between department managers and retail associates
The work content developed was compared between Old and New store structure
PMI provided the data required to support a demand-driven scheduling system based on workload and performed several detailed analyses on this data. This data can be used to ensure the appropriate service level is achieved, without overstaffing. Several process improvements were suggested based on lean principles which will enable the retailer to improve productivity of staff and to improve the customer’s shopping experience.
A simulation model was built to evaluate the performance of an automatic warehousing system under different values of design parameters and operational policies. The automatic warehousing system considered was composed of two main segments: the AS/RS and marshaling system. The AS/RS included the stacker cranes, storage racks (bins), input buffers, and output buffers. The marshaling system included the output dock, input dock, pick loop, and the input/output conveyor. The two systems interfaced at the input and output buffers of AS/RS. The performance of the warehousing system was observed under a large number of design parameters and operational policies including the number of aisles in AS/RS, the number of horizontal and vertical bins in each aisle, aisle assignment policies, bin assignment rules, and conveyor and stacker crane speeds.
The incoming parts are first loaded to (input) pallets and the contents of each input pallet are communicated to the central computer. The computer assigns an aisle number and a bin (rack) location to each input pallet. The input pallets queue at the input dock. Whenever an empty place is detected on the conveyor, the input pallet is transferred to the conveyor. When the input pallet reaches its designated aisle, the computer checks for an empty place in the input buffer of the aisle. If an empty place in detected, the input pallet is automatically transferred to the buffer. Otherwise, the input pallet makes one complete cycle and tries again to enter to its designated aisle. The automatic stacker crane of the aisle finally picks the pallet from the input buffer and stores it to its assigned bin. When an item is requested from the system, the computer selects the aisle and the rack location of the (output) pallet. The stacker crane picks the output pallet from its bin and transfers it to the output buffer. Whenever an empty place is found on the conveyor, the output pallet is moved to the conveyor and transported to the output dock. The output pallet may then enter a pick loop or it may be emptied and returned to the incoming stock location for reuse. The pallets that enter the pick loop are sent back to the AS/RS after their contents are altered.
The client, a major automotive manufacturer, wanted to verify the throughput of the system under different conditions. The performance of the warehousing system as demand (store and retrieve requests) doubled and tripled was to be observed. The number of aisles (stacker cranes), the capacity of each aisle, and the rules for selection of the store and retrieve requests for processing by the system had to be decided.
The overall objective was to design an automatic warehousing system that was efficient and flexible. The best values of a number of system variables and operational policies had to be decided which included:
Number of aisles
Vertical positions of input and output buffers of each aisle
Capacity of input and output buffers
Number of vertical and horizontal bins at each aisle
Policy for storing different part types (based on part turnover frequency) at the aisles
Policy for storing different part types (based on part bin size requirements) at the aisles
Horizontal and vertical maximum/minimum stacker crane speeds
Store and retrieve request profiles for each part type
Bin assignment rules for store and retrieve requests (closest – open location rule, first-stored, first retrieved rule, closest – full location rule, frequency class-based closest – open location rule)
Position of input and output docks.
The results of the simulation determined that:
Crane speeds can affect the throughput rate of the system by 30 percent.
Closest – full location bin assignment rule increases throughput by 10 percent when compared to first-stored, first-retrieved bin assignment rule for the retrieve requests.
A 3 – aisle system degrades the throughput of the system by about 10 percent when compared to a 4 – aisle system.
The proposed 4 – aisle system can perform under acceptable conditions even if the demand for
the system increases by three hundred percent.