Sunteți pe pagina 1din 6

IE361 Operations Research III Case Study 2 Middle East Technical University, Fall 2012 Due Date: January

21, 2012

Line Balancing Assembly lines are flow-oriented production systems which are still typical in the industrial production of high quantity standardized commodities and even gain importance in low volume production of customized products. An assembly line consists of (work) stations arranged along a conveyor belt or a

similar mechanical material handling equipment. The workpieces (jobs) are consecutively launched down the line and are moved from station to station. At each station, certain tasks are repeatedly performed in a limited time called the cycle time. An indivisible portion of the total work content in an assembly process is called a task, and the necessary time to perform a task is called the processing time (that can be deterministic or stochastic). The tasks are allocated to stations according to given precedence relationships among them. After a cycle time, the unfinished products are moved from one station to their successive station until they reach the end of the line. At the last station, the (end) product is completed and then it leaves the line. The decision problem of optimally partitioning (balancing) the assembly work among the stations with respect to some objecive is known as the Assembly Line Balancing Problem (ALBP). The data elements of simple ALBP are the set of tasks, processing times and precedence relations of these tasks. An example ALBP is given in Table 1 and visualized by the precedence diagram in Figure 1.

Figure 1. Precedence diagram of the example ALBP 1

Table 1. Data of the example ALBP Task 1 2 3 4 5 6 7 8 9 10 Immediate Predecessor(s) 1 3 1,4 5 2 6,7 8 9 Processing Time 6 6 5 5 4 5 4 2 9 2

Consider task 8 in order to illustrate precedence relations. Direct predecessors of task 8 are 6 and 7. By backtracking the direct predecessors, one can obtain the set of all predecessors of task 8 as {1,2,3,4,5,6,7}. This relation can also be observed from the precedence diagram. Then, task 8 cannot be started until all its predecessor tasks are finished. In other words, task 8 cannot be assigned to a workstation which is earlier than any of the workstations that its predecessors are assigned to. In Figure 2, three different assignments are displayed. For each assignment, the columns correspond to the workstations denoted by bold figures. For example, there are three workstations in assignment (i): tasks 1, 3 and 4 are assigned to workstation 1, tasks 5, 6 and 8 are assigned to workstation 2, tasks 2, 7, 9 and 10 are assigned to workstation 3. Note that the assignment in (i) is infeasible since task 8 is assigned to a station earlier than task 2, which is a predecessor of task 8. In (ii), a feasible solution with three workstations is given. The sum of processing times of the tasks assigned to stations 1, 2 and 3 are 16, 15 and 17, respectively. Then, the cycle time of this assembly line is 17. On the other hand, the assignment in (iii) constructs a longer assembly line with five workstations. Total processing times assigned to these workstations are 11, 11, 9, 6 and 11, respectively. Therefore, this assembly line operates with a shorter cycle time of 11.

1 1 3 4

2 5 6 8

3 2 7 9 10

1 1 3 4

2 2 5 6

3 7 8 9 10

1 1 3

2 2 4

3 5 6

4 7 8

5 9 10

(i) Infeasible Assignment

(ii) Feasible Assignment 1

(iii) Feasible Assignment 2

Figure 2. Different assignments for the example ALBP

There are two classical objectives of the ALBP: (1) minimizing the number of workstations (given the cycle time) and (2) minimizing the cycle time (given the number of workstations). All tasks assigned to each workstation should be finished before the cycle time. Therefore, the cycle time actually determines the production rate. Given a target production rate, and hence, the cycle time, the first objective is to minimize the number of workstations so that the required manpower, the length of transportation system and the required area for the assembly line can be minimized. On the other hand, given an assembly line and manpower, i.e., number of workstations, the second objective tries to minimize the cycle time so that the production rate is maximized.

References Roshania A., Fattahib P., Roshanic A., Salehid M., Roshani A. (2012). Cost-oriented two-sided assembly line balancing problem: A simulated annealing approach. International Journal of Computer Integrated Manufacturing, 25, 689715. Becker C., Scholl A. (2006). A survey on problems and methods in generalized assembly line balancing. European Journal of Operational Research, 168, 694715.

Case Study For the situations and objectives in parts I and II below, use Dynamic Programming to obtain the optimal solutions. In your work, please provide stage and state definitions, decision variables and recursive functions, explicitly.

Hint. You may consider the following as a state description: subsets of tasks which do not violate the precedence relations. For instance, if implies that all the predecessors of task is considered for a state, then

are also in . (Note that there exist 49 such

subsets for the numerical problem given in this case study.)

For each question in each part, you are supposed to give not only the optimal cycle time / optimal number of stations, but also the optimal assignment of tasks to the workstations.

Bonus. If your code does not only give optimal cycle time / optimal number of stations, but also gives the optimal assignments of the tasks to the stations, you will get extra 20 points. (Otherwise, you should obtain the optimal solution by backtracking manually and report the optimal assignments.)

You are supposed to submit the hard copies of your reports to assistants and upload your codes to online.metu.edu.tr. Please give all your documents in one file and upload it as a compressed file. Name the file with the names of your group members. Please name all your Excel and Matlab (or other) files properly, and in the appropriate parts of your report refer to these files.

PART I: Deterministic Case Suppose the processing times of the tasks are known with certainty as in Table 2. Assign the tasks to workstations to minimize the cycle time. Assume that you have solve the problem for every possible value of . workstations, and

Table 2. Immediate predecessors and processing times of tasks Task 1 2 3 4 5 6 7 8 9 Immediate Predecessor(s) 1 2 2,3 4,5 5 6 Processing Time 6 2 5 2 6 2 5 7 2

PART II: Stochastic Case Suppose that the processing times of the tasks are normally distributed random variables with expectations and standard deviations given in Table 3. Note that in this case, the cycle time ( ) is a random variable. Assuming that the performance times of the tasks are independent, use your statistics knowledge to comment on the distribution of the cycle time.

Table 3. Immediate predecessors, expectations/standard deviations of processing times of tasks Processing Time Task 1 2 3 4 5 6 7 8 9 Immediate Predecessor(s) 1 2 2,3 4,5 5 6 Expectation 6 2 5 2 6 2 5 7 2 Standard Deviation 0.4 1 0.6 0.9 0.8 1.2 1 0.8 1

Let the confidence level

be the minimum probability that the total processing times of , does not exceed some critical cycle time . That is,

the tasks assigned to a workstation j, say ( )

for all . The critical cycle time is specified to achieve a desired output rate

coming off the assembly line to ensure on-time delivery and also to prevent build-up of unwanted inventory. Hint: Let ( ) where denotes the cumulative distibution of the Standard Normal ( ) , the following inequality should be

distribution. Then, in order to satisfy satisfied: [ ] ( ).

a) What is the lower bound for b) Given that you have c) Given

) when

for all ?

, assign the tasks to workstations to minimize the cycle time. Assume workstations and solve the problem for every possible value of . and time units, assign the tasks to workstations to minimize the

number of workstations. 6

S-ar putea să vă placă și