Small Lot Delivery System

Transcription

1 Capstone Project Final Report Small Lot Delivery System IE 4800/4880 Fall 2008/Winter 2009 Team Members: David Naujokas (Project Leader) (734) Nashwan Fatteh (Team Member) (313) Instructor: Dean Pichette Alper Murat, Ph. D amurat@wayne.edu Office: Cell:

2 Acknowledgments We would like to give a special thank you to our Wayne State University Senior Project advisors, Dr. Alper Murat and Dr. Dean Pichette for their help and support in this project. Also Dr. Darin Ellis who helped the group in the ergonomics part of the project. Executive Summary Currently, Chrysler is using excel spreadsheets to do their efficiency and ergonomic calculations. The problem that was given was to produce AMPS, Automatic Manufacturing Planning System, screenshots that will allow Chrysler to do their calculations within AMPS instead of the excel spreadsheets and manual methods that they are using now. Before the screens could be produced, the efficiency spreadsheet had to be analyzed to determine where the numbers were coming from and to make sure that the way it was being done was correct. For the ergonomic assessment, the equations had to be found and the appropriate data needed to be in the spreadsheet. The equations that were used were the NIOSH 81 multi-lift equation and NIOSH 91 multi-task equation. With the ergonomic advancements, the IE will not have to leave their desk in order to do this assessment. Before the OR model (Operations Research), the tugger driver would complain that their list gave them problems then the IE would take that list and run the ergo assessment by picking the parts that were on the list in AMPS and run the assessment. With the OR model, the ergonomic assessment is in place while calculating the routes and hours for the day. 2

3 Table of Contents Acknowledgements...2 Executive Summary...2 Background...4 Previous Problem... 4 Problem Statement...5 Project Objectives...7 Methods & Approach...8 Constraints for Model...8 NISOH 91 Lifting Equation...9 Objective Function Model Creation Results and Expected Impact

4 Background AMPS is currently used as a user-friendly planning tool for direct-labor control and man-power assignments. Indirect labor (material handling) assignments for bulk materials are in the process of being integrated into AMPS. Small lot deliveries constitute a significant portion of the line feeding operations in Chrysler assembly plants. Further, its importance is expected to grow in accordance with inventory reduction efforts. Work assignments (Tugger operators, routes, delivery schedules) for small lot deliveries are currently performed via spreadsheets. Whenever there is a need for change in these assignments (i.e., efficiency, ergonomic considerations), these spreadsheets need to be manually updated. Previous Problem The challenge that Chrysler faced is that the indirect labor is not being implemented in the Automated Manufacturing Planning System (AMPS). The Small Lot Delivery System which is the indirect labor consist of Tugger routes is only being recorded and updated on an excel spreadsheet. The spreadsheet is a good process to monitor the routes, but employees manually updating the spreadsheet will cause data to be incorrect. Every time the spreadsheet gets updated there is no standard process that the employees follow. The problem with efficiency was it had to be validated by the group and make sure that the way efficiency was being determined was taking into account all the fixed variables. The ergonomic problems that Chrysler is having is the time that it takes time for the employees to go and study the routes of each individual tugger driver. They have to follow the driver and get the appropriate measurements in order to perform 4

5 the assessment. The specific problem to this is that tugger driver has a different route every day, so Chrysler is only getting the day s ergonomic assessment, and not the overall assessment of that route. So the problem is that Chrysler would like to implement these spreadsheets into their AMPS program. Problem Statement The challenge that Chrysler is currently facing is Routes are determined solely on hourly usage rather then a complete shift. The IE s would rather have a list for the tuggers on a daily basis then an hourly basis. This option will save time and money. Ergonomics is done after routes and efficiency is calculated. This is a problem because the tugger would complain of heavy lifting or too many parts. The IE would then have to collect data and measurements to check for any ergonomic concerns. All the data for the Tugger routes are in excel spreadsheet. This group wants to design an OR model that will help save the company time and money. The OR model will not only plan routes but will also include ergonomic feasibility checks. 5

6 Figure 1. Tugger Driver Process Tugger Picks up Pick List Tugger Picks up Parts in CMA (First Lift) Tugger Drives to Drop Location Is there still parts on tugger Tugger Drives to empty box location drop point Tugger drops off part at location Tugger Drops off empty boxes and Sorts them by size Tugger picks up empty container and places it on tugger. Tugger route is complete 6

7 Project Objectives Chrysler has asked WSU for help in fulfilling their need to put in place an OR model to plan routes. The OR model will also include ergonomic checks within the model. The project objectives were to analyze, understand and design an excel spreadsheet based work assignment for the Small Lot Delivery System which includes a segment of an assembly line. 7

8 Methods & Approach In the beginning stages of the project, the problem Chrysler is having with efficiency and ergonomics needed to be defined. The steps used for this project were: first, go through the Tote Simulation spreadsheet that was provided by Chrysler to clearly understand what is involved in creating spreadsheet; second, to find any mistakes that were made; third, determine what the necessary constraints are for a model; and fourth, to determine how to incorporate the NIOSH 91 equation into the model. The data was collected by the previous group that completed their capstone in 2007/2008. Due to the time it was very unfortunate that this group didn t have the chance to take a tour to the plant to getter a better understanding of how the tugger replenishes the lines. Constraints for Model There are three main constraints for this model. First, the inventory levels must be kept at a level sufficient to prevent any line stoppages from lack of parts. This requires a minimum constraint for each station greater than zero. Second, the inventory levels cannot exceed the actual capacity at each station. This provides a maximum inventory constraint for each station. Since actual maximum inventory values were unknown to the team, they were estimated based on part consumption rates. The second main constraint is that delivery amounts can only be integer values. Inventory only arrives in full containers and therefore, the tugger operators cannot move partial part containers to the assembly line. 8

9 The third main constraint is that relating to the ergonomics. The ergonomic equation used for this project is the NIOSH 91 lifting equation. The implementation of this equation is explained in the next section. NIOSH 91 Lifting Equation The NIOSH 91 lifting equation is as follows: RWL = LC x HM x VM x DM x AM x FM x CM RWL = Recommended Weight Limit LC = 51 lbs HM = Horizontal Multiplier - HM = 10/H, where H is the distance from the hands to the midpoint between the ankles. It can be estimated using the following two equations: When V is greater than or equal to 10, H = 8 + W/2 where W is the width of the container. When V is less than 10, H = 10 + W/2 If H is less than 10, the multiplier equals 1.0. VM = Vertical Multiplier VM = 1 ( V-30 ), where V is the vertical distance between the hands and the floor. DM = Distance Multiplier DM = ( 1.8/D ), where D is the vertical travel distance of the object. If D is less than 10 inches, DM equals 1.0. AM = Asymmetry Multiplier AM = 1 ( A ) The asymmetry angle, A, is measured from the mid-sagittal line of the worker. For the purpose of this project, the AM is assumed to equal 1.0 since all of the lifts are either from a general storage location or directly off of the tugger meaning that the container is directly in front of the worker when it is being lifted. 9

10 FM = Frequency Multiplier F is the frequency of lifts in lifts/min. It can very from 0.2 lifts/ min up to 15 or greater. There is no equation for this multiplier and must be pulled from a table. This table is shown in figure 2. CM = Coupling Multiplier CM is a classification modifier with three options, Good, Fair, and Poor. It is also based off of a table shown in figure 3. For this project, CM is assumed to be 1.0 since all of the containers have good handles and the components are unlikely to shift during the lift. Figure 2. Frequency Multiplier Table 10

11 Figure 3. Frequency Multiplier Table 11

12 Objective Function It was determined that there can be several objective functions for this model. It can be a. to minimize the total work of the tugger operator, b. to minimize inventory levels of parts on the line, or c. to minimize the cost both underutilizing the tugger operator and the cost of inventory levels. The equations for calculating total work were directly based on the efficiency calculations for the previous year s project. Model Creation The left side of the model is based directly upon the previous year s project. Figure 4 captures a screenshot of this section. Figure 5. Left and Middle Sections of Model Since the only variables for the NIOSH equation that are different for each part number are HM, VM, and L, these are the only multipliers displayed on the model. FM is also listed but is based on the hourly frequencies. In order to simplify the model, HM 12

13 and VM are chosen to be the smallest multiplier for which containers are chosen for each hour. In other words, the model chooses the smallest multiplier for which parts are actually being moved on the route. L is also chosen in the same way except that the largest one is chosen. This inherently adds a factor of safety for each route because the RWL will always be smaller than what it would be if a multi-lift equation was used. In order for the model to be expanded to include the full number of parts that are used in SHAP, the multi-lift equation could not be used due to its much higher complexity. The middle section of the model is where the solution and the RWL are displayed. The solution is broken into hour routes where the numbers represent the number of containers to be delivered to each station for that hour s route. Just below the solution are the NIOSH equation calculations. It shows the HM, VM, and FM used for each hour s equation. It then calculates a RWL for that route and checks it against the maximum L for that hour. The following screenshot shows the inventory calculations for each part. Figure 6. Right Side Section of Model The Max All Inventory column is the maximum allowable inventory for each part number. For the purpose of this project they are estimated based on the hourly 13

14 consumption of parts. The next column, Max Invent, is the actual maximum inventory level for each part. The Initial Supply column is where initial inventory levels before hour 1 are entered. Results and Expected Impact The model created for this project successfully finds solutions that meet the constraints established. However, the objective function of minimizing inventory failed to produce any results. This may be a limitation with the solver function in excel. On the other hand, however, setting the objective function to minimize the cost of both underutilization and cost of inventory successfully solved the model. This is almost the same as minimizing inventory except has the added benefit of maximizing the work of the tugger operator. It should also be noted that the model cannot find solutions when the initial inventory values are equal to zero. What this means is that there must be some inventory at the stations in order for this model to be applicable to the plant floor. The reason that it cannot find a solution is because the tugger operator would have to deliver parts to every single station if the initial inventory was zero and this would greatly violate ergonomic feasibilities. The impact of this project should be that Chrysler will be able to create routes for their tugger operators that will both be the most efficient as possible as well as keeping those routes ergonomically feasible without ever having to do an ergonomic assessment at the floor level. This will both save lots of time for both the industrial engineers and the tugger operators. They may also be able to further reduce costs due to lowering overall inventory levels. 14