SIX SIGMA PROJECT
improving TrainCo PERFORMANCE
GROUP #
Name Net ID
Name Net ID
Name Net ID
Name Net ID
Date
Summary
(Do this part at the end of the Six sigma project when you complete the Control stage) Explain company’s problem and it’s current situation. Explain how you applied Six Sigma to improve the problem. Share some details of your findings and how you could improve the situation in the company.
DEFINE PHASE
This project focuses on the improvement of TrainCo customer’s satisfaction level, specifically regarding the delays on arrival and departure of trains. This phase aims a better comprehension of the problem, setting target, scope process and considering voice of customer to seek the best possible approach to deliver a better result. The Define phase helps to have a better understanding of the problem, before investment on the project.
Project sponsor
TrainCo’s CEO
Project team
Process step
Represented by
From department
Role*
Can/Know/Care**
Planning
Rachel
Schedule Planning
Team Member
Know/Care
Maintenance
Chandler
Operations
Process Owner
Know/Care
Departing
Phoebe
Train operator
Team Members
Care
Monica
Train staff
Team Member
Care
Driving
Sheldon
Train operator
Team Members
Care
Stopping and walkout of the train
Theodore
Train staff
Team Member
Care
Overall Process
CEO
Board of Directors
Project Sponsor
Can/Care
Joseph
Swedish Railway System
External Data Source
Care
Name
Six Sigma Team
Project Manager
Know/Care
Name
Name
Name
Six Sigma Team
Six Sigma Temporary Consultants
Know/Care
Background and Qualitative Problem Statement
Explain, in a few lines, the company’s background and the problem that is brought to you based on the assignment document.
Defect Comments
Defect 1 = Arriving with delay of more than 15 minutes to the last station
Defect 2 = Earlier departure or departure with more than 15 minutes delay
What is the formula for Big Y2 based on the assignment document and the Defect 2 comment above?
Quantitative Problem Statement
Problem 1. According to customers’ opinion, more than half of the arriving trains to the last station have delays.
Problem 2. According to customer’s opinion, more than 30% of the trains depart from the stations earlier than the scheduled time.
Quantitative Goal Statement
More than 90% of the trains should arrive to the last station with less than 15 minutes delay. In other words:
Y1 ≥ …
More than 95% of trains should departure between the scheduled time and 15 minutes after the scheduled time. In other words:
Y2 ≥ …
Scope and Boundaries
Project Scope: Transferring Passengers between New York, NY and Schenectady, NY.
Out of scope: control of the weather, change the resources (trains, rails, train traffic, etc).
Within scope: Optimize the process with the resources already possessed i.e. trains, railways, train traffic and etc., scheduling, maintenance planning and human resource management.
Project boundaries:
Time: 4-09-2021 to 4-23-2021
Resource: Possessed trains
Information: Only the data got from TrainCo’s CEO
Deliverables
Improve client satisfaction, specifically towards the delay complaints and control the process between specification limits required by TrainCo’s CEO.
Less than 10% of the arriving trains to the end station will have delay more than 15 minutes.
No trains leave the stations before the scheduled time and less than 5% of the departing trains will have delay of more than 15 minutes.
Business Benefit Estimation
Skip this part, as not much information is given in the assignment instructions.
You can, however, use some arbitrary estimation; for example, based on the information of the assignment document we know that there has been 20% reduction in the sales of the tickets.
You are not given the total number of passengers per day for the trains, but you can assume a value.
E.g. before the reduction of ticket sales, let’s assume that each day on average the company sells 2000 tickets and the cost of each ticket is $50 on average. So the daily ticket sales is about (2000*$50 = ) $100,000. So, 20% of the total sales is $20,000 which is the daily loss of revenue per day for the company because of the delays.
If you can fix the delay issue and bring these passengers back to your company, you potentially can bring back $20,000 of revenue per day ($20,000 * 22 working days per month = $440,000 per month revenue); this could be the potential business benefit of this project.
In general, remember that most of the times it is possible to do a back-of-the-napkin calculation and estimate a monetary value as I did above.
Project Charter
Fill out the form of the project charter template that I have shared with you and copy it here. (You can copy an image of it to make things look cleaner). To fill different sections of the form, you should use the information given in the assignment document and also the define section of this document.
Over-all process map (SIPOC)
Voice of the Customer (VOC)
TYPE OF QUALITY REQUIREMENT
WHAT DOES THE CUSTOMER EXPECT?
WHO IS THE CUSTOMER?
delighters =
wishes and unknown delighters (surprises)
Place to sit
Free Wifi
High-quality materials inside the train
Train arriving ahead of schedule
Climatization of wagons (AC or heating)
Good money returning policy (refunding)
Passengers
Passengers
Passengers
Passengers
Passengers
Passengers
more is better =
specified demands and expectations
Wagon not too crowded (place to stand)
Train arriving within 8% delay (tolerance)
Value of money for the ticket
Cleanliness of the inside
Cleanliness of the outside
Passengers
Passenger, train staff and station staff
Passengers
Passengers and TrainCo (outsourcing)
Passengers and TrainCo (outsourcing)
must be =
absolute demands
Train in good condition to arrive at destination
Trains not cancelled
Presence of toilets inside the train
Space for the luggage
Passengers
Passengers and train staff
Passengers
Passengers and staff in general
Passengers
Critical to Quality (CTQ) Tree
Stakeholder Analysis
Assign (type) each of the following to the category that you think is the most appropriate. (this is an open question, so just trust your judgement about the assignments). I assigned a couple of them to give you the idea.
Government (Regulations)
TrainCo’s CEO
Six sigma team members
Department of Transportation
Media
Competitors
Passengers
Weather Forecasting Agency
Shareholders
Employees of TrainCo
Government (Regulations)
TrainCo’s CEO
Measure Phase
The measure phase aims to set a stake in the ground in terms of process performance through the development of clear and meaningful measurement systems. Based on the data received for the travel time fluctuations between different stations against all the trains operated under TrainCo., a measurement system has been prepared using certain standard statistical tools to measure the capabilities and difficulties in the process which ultimately leads to better understanding of the problem and identification of the root cause.
Measurement variables
Data Collection Plan
In general, having a data collection plan is very helpful as you know what type of data you exactly need in order to move toward the goals of the project.
Train delay
Train delay
Arrival delay at final station
Arrival delay at final station
Departure delay at all stations
Departure delay at all stations
Average delay
Average delay
% of trains with > 15 min delay
% of trains with > 15 min delay
Average delay
Average delay
% of trains with > 15 min delay
% of trains with > 15 min delay
% of trains with early departure
% of trains with early departure
KPI1
KPI1
KPI2
KPI2
KPI3
KPI3
KPI5
KPI5
KPI4
KPI4
KPI
Average delay of trains in arriving at the final station
Percentage of train arriving at the final station with more than 15 minutes
Average delay in departure at all the stations
Percentage of trains departing earlier than the scheduled time
Percentage of trains departing with more than 15 minute delay
Sampling
Read this part carefully to understand the data
Sample Number: (Numerical Data) provides the number of samples
Date: (Contextual Data) shows the travel date of each train. From 01.01.2020 till 03.31.2020
Train No.: (Contextual Data) Four Train numbers are available SN-1,2 and NS-1,2. Where SN is the abbreviation for trains travelling from Schenectady (S) to NYC (N) while NS represents the trains travelling from NYC (N) to Schenectady (S). 1 and 2 reflects the scheduled departure at the starting station. The trains departing from the first station at 6 ‘o’ clock are named as 1 and those starting their journey at 16 ‘o’ clock are termed as 2.
Crew: (Contextual Data) Total 6 crew teams are present. They are further divided into two more groups N (N1, N2, N3) and S (S1,S2, S3) based on the location, where N means NYC and S means Schenectady. The crew consists of one driver and two train hosts.
Station: (Contextual Data) Schenectady, Albany, Hudson, Poughkeepsie, Yonkers Station, and NYC- these six stations have been considered and observed for data collection and analysis.
Arr/Dep: (Contextual Data) it represents the status of the train: departure, Arrival at each station and Arrival at end station.
Schedule: (Numerical Data) provides the scheduled time (expected time) of the trains against departure, Arrival and Arrival at end station.
Actual: (Numerical Data) provides the actual (practical) time of departure, arrival at each station and arrival at end station.
Added Time: (Numerical Data) this is the individual contribution of each station to the overall delay when comparing the arrival and departure at the same station or departure at one station and arrival at the next station.
Total delay: (Numerical Data) it is the difference between Actual time and Scheduled time against each departure, arrival and arrival end.
Descriptive statistics
Pivot Table
Task 1-1) Create a pivot table that summarizes the average of “Total Delay” for each station. The pivot table should look like the following.
Boxplots
Example) Create a box plot that is comparing different stations total delay for arrival. To do this, first filter column G (Arr/Dep) by selecting only “Arrival” and “Arrival End”. Check the uploaded video as a guide. Copy the chart here. To copy a chart from Excel to Word, make sure you paste a picture of the chart.
Task 2-1) Create a box plot that is comparing different stations “total delay” for departure. To do this, first filter column G (Arr/Dep) by selecting only “Departure”. Copy the chart here. To copy a chart from Excel to Word, make sure you paste a picture of the chart.
Explain what you understand about the chart.
Task 2-2) Create a box plot that is comparing different stations “Added Time” for departure. To do this, first filter column G (Arr/Dep) by selecting only “Departure”. Copy the chart here. To copy a chart from Excel to Word, make sure you paste a picture of the chart.
Explain what you understand about the chart.
Control Charts
Task 3-1) Create an I-MR chart of “Added Time” for the departures. Calculate control limits for the I-MR charts and use Excel to create the charts. Watch the recording of Week 5 to learn how you can do this. Copy the chart here. To copy a chart from Excel to Word, make sure you paste a picture of the chart.
Explain what you understand about the chart. Is the departure process in control?
Problems/issues Identified in Data
Summarize what you understand from the pivot table and the charts that you created? Are delays more common in any specific station(s)?
Measured Big Ys
Use the data to calculate Big Y1 and Big Y2 based on their formulas that you have from the Define stage.
Big Y1 Current yield = ?%
Big Y2 Current yield = ?%
AnalyZE Phase
(For the Analyze and Improve phase, you are not given any tasks due to limited available time in our course. I encourage you to have a look on some of the tasks that are done in these phases. Note that this is just a summary of what can be done for a real project. Keep this in mind that each of these phases requires detailed attention and time to be completed successfully. You can reach out to me if you have any questions about these sections.)
The Analyze Phase aims to analyze and verify the considered root causes to determine if those issues are the real root causes of the problems experienced by TrainCo. For this section, two approaches are applied:
Data Door approach
The results from the Measure section are used to find any significant data, i.e., something uncommonly different from the data collected.
Process Door approach
Using 5whys and the Fishbone diagram, the process steps causing problems were identified and the verified.
Analyze part I: Find patterns in data
Data door analysis
Hypothesis tests
One-way ANOVA: Added time versus Station
In the Null Hypothesis, we assume that the mean of added times for all stations are equal.
H0 : µNYC = µYonkers = µPoughkeepsie = µHudson = µAlbany = µSchenectady
H1 : At least one of the means is different from the others
Source DF SS MS F P
Station 5 31060,97 6212,19 684,22 0,000
Error 1794 16288,20 9,08
Total 1799 47349,17
S = 3,013 R-Sq = 65,60% R-Sq(adj) = 65,50%
From the chart above, according to the low P-value, we can reject the null hypothesis, so there is significant difference between the results of the stations. This difference is caused by Hudson which is the outliner station and this also shows that is very unlikely that the differences in data are due to random variation.
Analyze part II: Find root causes
Process door analysis
Cause and effect diagram has been constructed for both our specific cause variable ‘Departing late at Hudson station’ and general cause variable ‘Travel time variation’. The Ishikawa uses 7M reasoning (Man, Machine, Methods, Milieu, management and materials. Within each M, we used ‘why why’ analysis at least for 5 times to reach to the root cause. E.g- When analyzing for special cause variable the 5 why analysis was adopted in the following manner:
1. Why? – Problems exist in the external environment
2. Why? – There might be problem in the Hudson platform
3. Why? – The platform gap is larger.
4. Why? – The platform is old and not good for the new trains
5. Why? – Extra arrangement required for exit and boarding procedure (time consuming)
Each person in the team identified the root cause, which he believes, to contribute most heavily towards the problems faced by TrainCo. After having a detailed analysis of the system with respect to both special and common causes, we found inefficient planning and communication mismatch (both within the crew and operation team) as some of the most vital root causes.
Analyze part III: Verify root causes
Verification of Root Causes
The root causes then, were first divided in to ‘in scope’ and ‘out of scope’ based on the question whether “would we be able to improve them or not?” The ‘in scope’ root causes were then further divided into 5 groups according to the area of origin, out of which ‘Insufficient Planning’, ‘Insufficient Communication’ to be the most contributing factors. For the verification of the root causes, the information was shared with the representatives of TrainCo. To set the improvement priorities, the brainstorm results of group names were put in the Effort-effect diagram and they are represented by the following figure. As can be seen, inefficient planning and communication are prioritized.
EFFECT
(on Big Y)
EFFECT
(on Big Y)
Improve Phase
The Improve Phase aims to plan how to implement the solutions prevenient from previous activities. Here, starting with the Action Plan we generated solutions and evaluated and selected then to develop a Risk Analysis/FMEA later on and then to execute a pilot test for the process improvements. Finally, the implementation plan is presented in the Gantt chart.
Action Plan
Verified root causes are divided in two groups:
Insufficient planning:
Management > bad planning, scheduling > overlapping with other train companies > no coordination with Department of Transportation (DoT) Traffic Control
Milieu > Late Signal to Depart > Schedule overlapping > Poor Planning
Inefficient communication:
Machine > Communication mismatch > Insufficient IT, Software and devices at platform
Generation, synthesis, and selection of Solutions
Solution generation
In order to come up with solutions, we have used the following techniques:
Benchmarking: The present practices adopted by TrainCo have been compared with the other railway companies operating across the globe. The comparison has been included in the ‘comment section’ of the table.
Negative brain storming: The problem was first taken as the reference. A negative approach was followed in order to know how to make the situation worse and to find out different root causes which might contribute to the worsening of such situations – e.g., Insufficient Planning was taken as the reference. We did brainstorming to find out how to make it worse and overlapping with the trains was found to be a good way to make the present planning worse. Then again after negative brainstorming, we reached to the conclusion that schedule of the trains operated by the competitors of TrainCo (such as Amtrak) can cause the overlapping.
Different Hats: In this technique, an imaginary situation was assumed if the team members were given different roles such as ‘representative of Department of Transportation’, ‘Driver of the train’, ‘crew’ and representative from train platform’.
Assumption busting: The proposed solutions were drawn questioning traditional assumptions in order to find out alternative processes against the old ones. For example, ‘Why does it have to be like that?’ and ‘Why can’t we do it differently?’ were common questions for this brainstorming tool.
No.
Verified root causes
Solution(s)
Solution Generation Method
Comment
1
Insufficient Planning
Coordination with DoT traffic control
Assumption busting
Why is the planning bad? Who has a relevant influence in the results?/Who should collaborate with us in order to result in a better planning?
Rescheduling using more advanced scheduling software
Benchmarking
‘Hastus-Train’ is advanced software used by both United Kingdom and Belgium Government for optimization of Train scheduling and railway planning in order to accommodate both minor and major changes in train schedules.
More efficient communication with DoT traffic control in order to be more flexible in facing changes/new regulations/etc.
Different Hats
Using this approach for TrainCo and DoT’s roles, it was clear to understand the importance of a fast and accurate communication between these two for basic operations and planning.
More efficient communication with station in order to be more flexible in facing changes/new regulations/etc.
Different Hats
Using this approach for TrainCo management team and the stations, the importance of an efficient communication between these two was set as a ‘must have’ for daily operations.
Better connection with competitors in order to coordinate the schedules with them
Negative brain storming
Bad planning > No collaboration/connection with competitors > Overlapping schedule with them.
2
Inefficient Communication
Upgrade IT devices
Benchmarking
Network Railway Company from U.K has improved communication efficiency and trust using GSM-R as a Railway Communication System.
Contract with a better internet/intranet service provider
Benchmarking
The trains in Ukraine have implemented ‘WiMax service’ which can deliver the Wi-Fi signals to all the trains operating within a distance of 1350km.
After potential solutions are generated and best solutions are selected, detailed risk analysis is performed to have a better understanding of the tasks that need to be done. Later Pilot studies are done to verify the effectiveness of the solutions in small scale. Pilot studies were successful. These parts are not presented in this document.
Implementation plan
The following implementation plan suggests a plan for the implementation of the improvement phase. As can be seen in the Gantt chart, it takes roughly two and a half months to finish the plan, in which forty days are for the rescheduling part. Activities 5, 6, 9, 11, 12 and 23 are located in the critical path, which is shown in red. (This is just an example for your information)
Control Phase
The Control phase deals with the improvement of the process based on the new data collected after improvements. The set of solutions has been implemented and the results have been compared with the original process to find out the improvements in TrainCo.
Standardization and education
As mentioned before, our main improvements are coordinating with DoT Traffic Control, upgrading IT devices and rescheduling using more advanced software. So, to keep the coordination with DoT Traffic system and competitors, periodic meetings should be set and the contracts should be updated to new situations. Besides, to standardize the new IT systems, we need to educate TrainCo staff to adapt them. In addition, they should become familiar with how to do the maintenance and backup the new central system. Thus, an education plan is designed as follows:
Training Plan
Purpose for Training
Topics Covered
Training Given by
Participants
Date
Coordination with DoT traffic control
Communication techniques with central agency during normal operations, how to coordinate.
Project Manager, DoT traffic control
Control engineers and operators
MM/DD/YYYY
Communication techniques during emergency conditions, sudden schedule change
MM/DD/YYYY
Platform Management, communication with platform staff
Platform Engineer
MM/DD/YYYY
Signal Processing
Instrumentation and Control Dept., DoT Traffic Control
Control engineers
MM/DD/YYYY
Upgrading of IT devices
Detailed technical specifications, Use, Maintenance
Technical Experts from IT Service Provider (Vendor)
IT Department
MM/DD/YYYY
Communication methods over the new devices from operations point of view, advantages and precautions
Operation Team (Crew, driver)
MM/DD/YYYY
Control and logic checking using new devices, communication patterns
Control engineers
MM/DD/YYYY
Normal operations and handling
All other employees
MM/DD/YYYY
Rescheduling Software
Accommodate changes according to requirement, Maintaining security, encountering problems maintenance issues
Software Developer
IT Department
MM/DD/YYYY
Detailed training on uses of software, benefits over the previous software, response to schedule defects raised by operation team and Analyzing of outputs with respect to input ranges
Planning Department& IT Dept.
MM/DD/YYYY
Raise scheduling defects if found any, to check the schedules using the software.
Operation department
MM/DD/YYYY
Monitoring and Control charts
Control charts
Task 1-1) Create an I-MR control charts that show before and after situations and highlights the improvements in added time of departures. Use Excel to create the charts. The before-after improvement control chart should look like the following example (the chart below uses some random data so your chart will look different; remember that you need both I-chart and MR-chart). You need to combine the old and new data to create these charts. Follow the instructions from the lectures to create these control charts in Excel.
Explain what you understand from these charts. Is the process in control after improvements?
Box Plots
Task 2-1) Create a box plot that compares the total delay for arrival of old data and new data for the end stations (NYC and Schenectady). The box plot should look like the following example (the chart below uses some random data so your chart will look different) . Use Excel to create the chart and copy it here.
Explain what you understand from these charts.
Task 2-2) Create a Box plot that compares the added times of departure for old and new data by station. Your chart should be similar to the following chart. (the chart below uses some random data for only 4 stations, so your chart will look different and includes all 6 stations). Use Excel to create the chart and copy it here.
Explain what you understand from these charts.
Verified process performance
Calculate Big Y1 and Big Y2 based on the new data
Big Y1 yield after improvements = ?%
Big Y2 yield after improvements = ?%
Did the project reach its goals that were set in the define phase?
Note that if the answer is Yes, everything is great; but if not, it is also fine. We can launch future projects to improve the situation further and reach our goal.
Project closure
Project Closure and Handover Sheet
Project title
…….
Business Benefit
….
Currency
USD
Project closure checklist
* Project results have been reported to the Process Owner and Sponsor.
* Deliverables required have been accomplished
* Goals outlined in the project charter have been achieved.
* Business case is confirmed
* Improvements are handed over to the Process Owner who is from now on responsible for applying the improvements and their controls established to ensure their sustainability and a process in control
Key roles
Name
Signature
Date
Process owner
TrainCo CEO
Project leader
Team member 1 name
Business Controller
Other team member’s names
Sponsor
TrainCo CEO
Go back to page 2 and write 1-2 paragraphs summary of the project!
Congratulations! Hopefully now you have a much clearer understanding about Six Sigma’s DMAIC.
66
