Optimizing Rolling Stock Fleet: Analysis and Recommendations

1

2 Introduction This report will cover the analysis of “Rolling stock fleet” by using tools such as tableau for graphical visualization and using python for predictive model. After successfully analyzing what is provided in the dataset, we can provide recommendations to the project executives to make inform decisions. This report is divided into three parts Tableau Visualization, Predictive Model and then providing Recommendations. Recommendations will include strategies to optimize, maintenance schedules, and improve operational efficiencies. Each process will provide a comprehensive understanding of the rolling stock fleet, which will empower the decision makers with information to effectively manage the fleet.

3 Data Analysis and Observation Figure 1 Analysis of Downtime and Labor hours by Equipment category The bar chart shows downtime and labour hours by equipment type: heavy equipment, heavy trucks, and light duty vehicles.  Heavy Equipment: A 1:1 ratio of 105,288 downtime/labor hours indicates reactive maintenance using labor only when downtime occurs. Lower values could result from improved dependability or a smaller fleet.  Heavy Vehicle: 185,026 hours suggests that increased downtime may have an impact on operational efficiency. Equal labor hours indicate a resource strain or increased maintenance requirements.

Your preview ends here

Eager to read complete document? Join bartleby learn and gain access to the full version

Access to all documents
Unlimited textbook solutions
24/7 expert homework help

4  Light-duty Vehicles: The most hours at 208,598 indicate potential maintenance issues or increased usage. Because of their critical role in operations or specific deployment, matching labor hours implies prompt care. Figure 2 Analysis of Equipment class with respect to Mileage The largest rectangle is light pickup vehicles, with 590,216 miles. Due to its mobility in a variety of terrains, light pickup trucks may be widely used in the fleet. Heavy pickup vehicles

5 with 538,454 miles follow. Heavy-duty trucks can handle harsher jobs and situations due to their durability. Their high mileage may indicate use in more rigorous jobs that wear out lightweight automobiles faster. Figure 3 Count of cleaned rolling stock fleet with respect to service group  Engineering services: Has the most cars (807) that were cleaned, possibly due to a large fleet of specialized vehicles that require frequent cleaning.  Parks and Recreation: Cleaning 290 vehicles, indicating a sizable fleet for park and recreational maintenance. Emission Analysis By utilizing "Cleaned_Rolling_Stock_Fleet.csv" is a file that contains fleet information such as "Equipment Class," "Equipment Category," "Service Group," "Make," "Model," and

6 "Mileage." We determine departments need a reduction in emission and to achieve the target of 20% reduction. With the help of data mapping, we applied the average emission by vehicle category to the primary data. Using linear regression, we made predictions (Y = ['Emission']) using features such as ('Make', 'Year', 'Mileage2014 (km)'). The model achieved high performance, which can be verified by the R2 and MSE which were (R2 = 0.9998, MSE = 0.0042) respectively. Departments that had the highest cumulative emissions were  Park Board Planning and Operations: 105968  Real Estate & Facilities Management: 117,712  Engineering Streets Operations: 131,220 To achieve the target of 20% reduction the determined cut-down threshold is 26244. Recommendations  A proactive, preventative maintenance strategy may decrease downtime if labour hours are reactive to equipment failure.  The data supports prioritizing vehicle cleaning in engineering and parks & recreation services, which may be related to their job and public interaction. Clean cars indicate professionalism and attention to detail, which may affect service groups' public image.  The predictive model suggests focusing on the Park Board Planning and Operations, Real Estate & Facilities Management, and Engineering Streets Operations. These departments have the highest cumulative emissions.

Your preview ends here

Eager to read complete document? Join bartleby learn and gain access to the full version

Access to all documents
Unlimited textbook solutions
24/7 expert homework help

7  We can see that the departments mentioned by the predictive model need most attention to reduce emission focusing on these departments and developing specific emission reduction policy in accordance with their usage patterns.  The benefits of reducing fleet size make way for operational flexibility, which means that a smaller fleet will be more agile in responding to changing demands in today’s world. It also helps in environmental sustainability since a reduced fleet corresponds to lower carbon emissions. This aligns with organization’s commitment to reducing its carbon footprint as well.  We can also develop strategies like predictive maintenance models and cross- departmental collaboration. These models will help reduce unexpected breakdowns and increase overall fleet reliability, whereas collaboration can lead to more efficient use of fleet assets and ensure cost savings.  Some of the potential risks include operational disruptions, resistance from user groups, technological challenges, etc. We can mitigate these risks by implementing a phased approach to fleet reduction, ensuring that essential services are not compromised. Developing contingency plans is the key here. Engaging in transparent communication by highlighting the overall benefits of fleet reduction and involving stakeholders in this type of decision-making can help with the resistance from users unwilling to give up fleet assets. Reduction in fleet brings in the need for comprehensive training on new technologies for staff. Conducting pilot programs to identify potential technical issues before full implementation can aid in this process. Pilot program can offer a more efficient resource allocation (Brigden, 2023).

8  Continuously communicate with the public, addressing concerns and proving the positive aspects of fleet optimization (Jang & Park, 2022). Managing public perception is crucial for maintaining community support throughout the project.  And finally, document and share best practices learned throughout the project, facilitating knowledge transfer. Sharing insights improves future decision-making and implementation strategies (Calzon, 2023). Conclusion This analysis provided an insight into the rolling stock fleet, emphasizing what are the strategies used for the maintenance. And what are the emissions for each department and what departments to prioritize to reach our target. Proactive maintenance, fleet optimization, and cross-departmental collaboration all contribute to increased reliability and sustainability. Risk reduction through phased implementation, stakeholder engagement, and open communication promotes successful adoption. References Brigden, B. (2023). Pilot project excellence: A comprehensive how-to (+ an example) . https://www.teamwork.com/blog/pilot-project-example/ Calzon, B. (2023, February 14). The importance of data driven decision making for business . BI Blog | Data Visualization & Analytics Blog | Datapine. https://www.datapine.com/blog/data-driven-decision-making-in-businesses/ Environment and Climate Change Canada. (2020). Greenhouse Gas Emissions Performance for the 2018 Model Year Light-Duty Vehicle Fleet . Canada.ca. Retrieved December 3, 2023,

9 from https://www.canada.ca/content/dam/eccc/documents/pdf/cepa/ldv-ghg- performance-2018-eng.pdf Jang, E., & Park, D. (2022). How can community support be maintained during the COVID-19 pandemic? Environment, Development and Sustainability . https://doi.org/10.1007/s10668-022-02765-x

Your preview ends here

Eager to read complete document? Join bartleby learn and gain access to the full version

Access to all documents
Unlimited textbook solutions
24/7 expert homework help

10 Appendix 1- Summary of analytics The data used for the analysis was “Cleaned_Rolling_Stock_Fleet.csv” which was cleaned of nun values. This dataset provided us with the information of the stock fleet with different parameters such as “Equipment Class”, “Equipment Category”, “Service Group”, “Make”, “Model”, “Mileage”, and many more. To check what which department needs to cut down on the emission we need to get the data for emission, there are only a few data for the emission for only selected companies, so we are going to use this data and then fill the average emission and map it to the primary data with this code. Appendix 1-figure 1 Code for mapping emission data “data['Emission'] = data['Make'].map(emission_standards).fillna(238)” This will map the emission data to the primary data and then we use that to predict the emission and decide what department is to reduce its emission to meet the target. Defining X as ['Make', 'Year', 'Mileage2014 (km)'] and Y as ['Emission']. We will train this data and use linear

11 regression model to predict further. Using pipeline, we can train the model to use the provided data and make new predictions on the unseen data. mse = mean_squared_error(y_test, y_pred) r2 = r2_score(y_test, y_pred) Using this code, we can check the model performance. The r2 of 0.9998049093787748 defines that the model is performing well as it closer the 1. The mse of 0.004272968300587835 explains further that the model is performing well as lower the mse better the model’s performance. For the prediction we need to set a target of 20% reduction in the emission for this we will use this code to set this target and define a code for the computation for cut down threshold. target_reduction_percentage = 20 cut_down_threshold = target_reduction_percentage / 100 * light_duty_data['Emission'].sum() The cut down threshold came to 26244, this is the value we need to reduce to get the target of 20% reduction in emissions. departments_to_cut = light_duty_data.groupby('Department') ['Emission'].sum().sort_values().cumsum() Using this formula, we can check the cumulative emission of deferent departments and check what departments concentrate on most. According to the results, park board planning and operations 105968.0

12 real estate & facilities management 117712.0 engineering streets oper. 131220.0 These departments produced the most emissions and needed to concentrate on reducing emission and meet the target of 20% emission reduction. To get the “Cleaned_Rolling_Stock_Fleet.csv” we used this code to clean the data before we went with the steps mention above: import pandas as pd # Load the dataset file_path = '/mnt/data/FA1 Scenario Dataset - Rolling Stock Fleet.csv' fleet_data = pd.read_csv(file_path) # Fill missing values with the median for specific columns fleet_data['Mileage2014 (km)'].fillna(fleet_data['Mileage2014 (km)'].median(), inplace=True) fleet_data['DowntimeHours2014 (hours)'].fillna(fleet_data['DowntimeHours2014 (hours)'].median(), inplace=True) fleet_data['LaborHours2014 (hours)'].fillna(fleet_data['LaborHours2014 (hours)'].median(), inplace=True)

Your preview ends here

Eager to read complete document? Join bartleby learn and gain access to the full version

Access to all documents
Unlimited textbook solutions
24/7 expert homework help

13 # For 'HourMeter2014 (hours)', fill missing values with the median and correct negative values fleet_data['HourMeter2014 (hours)'].fillna(fleet_data['HourMeter2014 (hours)'].median(), inplace=True) fleet_data['HourMeter2014 (hours)'] = fleet_data['HourMeter2014 (hours)'].abs() # Assessing and correcting the 'Year' column mode_year = fleet_data['Year'].mode()[0] fleet_data['Year'] = fleet_data['Year'].replace({1940: mode_year}) # Save the cleaned data to a new CSV file cleaned_file_path = '/mnt/data/Cleaned_Rolling_Stock_Fleet.csv' fleet_data.to_csv(cleaned_file_path, index=False)

14

Final Group project, predictive (1)

Related Documents