Final Engineering Thesis Project for Industrial Engineering. Design of a lean system in order to improve the efficiency of the turbo machinery department.
Flow and Throughput Improvement at Pratt and Whitney Canada
1. Flow and Throughput Improvement
Liat Artzy-Bush Amir Hanna Pavlos Lazos Ramon Saviñon Anthony Toteda Cynthia Volpe
PROJECT SCOPE DECISION SUPPORT TOOL
Problem Statement Objective
• Growing list of incoming design jobs and a fixed employee capacity • Provide a tool to assist the manager in selecting design jobs
• Lack of prioritization, as well as problematic resource allocation and scheduling, have become evident
Method
bottlenecks which are decreasing flow and paralyzing throughput
• User assigns a weight to each decision criteria, in order to
Results
RECOMMENDATIONS
rank jobs against one another, producing a prioritized list
Prioritized
Job List
Time Type Strategic
Benefit DST Features
Short Term 1 - Enter New Job
Date Date Due Job class Customer
Received Relationship •Payback
Period 2 - Calculate Rank
Objective • Health &
Safety 3 - Search Job
•Environmental
• Improve information flow, project prioritization and Impact 4 - Output List
5 - Release Job
distribution through resource allocation and scheduling
• Use DST before the module in order to prioritize design jobs
• Apply LP model to DST output in order to allocate jobs to workers
SIMULATION • Pull jobs in the following sequence: H-H-H-M-M-L
LP MODEL
Objective Objective
• Identify and validate an optimal job pulling sequence while restricting the load • Assign a maximum of 3 jobs per designer • Improve job scheduling and resource allocation
on each worker Method
• Maximize the amount of jobs completed
Method
• Test different scenarios of varying queue size and sequence Subject To:
• Measure system performance through total tardiness and throughput Long Term • Is the job assigned to a worker ?
Max Jobs
• Is the total available hours greater than the cumulative hours of the jobs assigned ?
Scenario Sequence
per worker
1 3-2-1 1 • Is every job only assigned to one worker ?
2 3-2-1 2
3
4
3-2-1
3-2-1
3
4
• Is at least one job assigned to each worker ?
5 3-2-1 5
6 None 1 JOB ID • Does every job begin and end before the required date ?
7 None 2 E7799
8 None 3 JOB ID
E5555
9 None 4
10 None 5
Next Job Enters NO JOB ID JOB ID JOB ID JOB ID
E3576 E3294 E7889 E2235 Available Hours
Parameter Name Distribution Used
Job Category Job Percentile JOB ID
EXPO(0.6) E4566 0 Worker 1 1888
Arrival Rate High (H) > 95%
Days
5 + WEIB(363, 0.515) JOB ID JOB ID JOB ID
Processing Time Medium (M) > 80% YES E1111 E4351 E3599 Available Hours
Hours
One working year
Replication Length Low (L) ≤ 80%
(262 Days) Worker 2 1888
0