Hiryuu Project Management

Home

About Us

Project Overview

Findings

Project Management

Documentations

Overview

Scope of Work

1. Data Gathering and Scoping
Data have been gathered from our sponsor which consists of data across 3 months and spanning across several countries. The data given to us are all in .csv formats.
2. Research on Software and Proposal Preparation
Our team has researched and discussed on the software and applications used in analysing and visualising the data with our supervisor and sponsor.

We will mainly utilise JMP Pro for the clustering, exploratory, and seasonality analysis and QGIS for our geospatial analysis. Our sponsor has expressed their preference for the final product to be in Power BI as they are familiar with it. In addition, they have expressed interest for the final product to be dynamic, meaning any input of future data can tap on the methods used to produce the results.

Currently we will focus on analysis on JMP Pro and QGIS and have plans to decide how to compromise the dynamic capabilities within our limitations.

3. Data Cleaning and restructuring

After looking and familiarising ourselves with the data, we discovered the complexity of the data. The data is not only large in size, contains a wide variety, and also inconsistent across different countries. This inconsistency is due to the differences in language used for reporting as well as different systems used. As such, certain coding languages could not be used to process the values and other solutions have to be looked into.

Due to the complexity and variety of the data, we will be conducting data cleaning by removing duplicates, ensuring consistency by inserting the header column, merging the data files, filtering the data by countries and checking that the data is appropriate and ready for analysis.

4. Data Modelling
1. Exploratory Analysis
2. Geospatial Analysis
3. Time Series Data Analysis
5. Dashboard Building

To offer a dynamic dashboard that can intake raw data from the providers and produce meaningful graphs and charts for the user to understand the logistics performance, we decided to utilise Rshiny. Both service providers engaged very different in their data collection, hence we decided to build 2 separate dashboards to handle the 2 very different formats. However, our delivery of analysis is similar for both dashboards. We were able to build the dashboard with a clearer end picture in mind as we had a c better understanding of the data after the earlier analysis done on JMP Pro.

Timeline

Tasks	Start	End	Teammates Involved	Status
Data Gathering and Scoping
Gather Data	Week 0	Week 0	All	Completed
Finalise Requirements with client	Week 0	Week 0	All	Completed
Scope Project	Week 0	Week 0	All	Completed
Research and Preparation
Explore software	Week 0	Week 0	All	Completed
Finalise Proposal	Week 0	Week 0	All	Completed
Create and update wiki page	Week 0	Week 0	All	Completed
Data Cleaning
Data collection	Week 1	Week 2	All	Completed
Data cleaning and restructuring	Week 1	Week 2	All	Completed
Resolve/remove incomplete data	Week 1	Week 2	All	Completed
Data Modelling
Stage 1: Exploratory Analysis	Week 3	Week 8	All	In Progress
Stage 2: Clustering	Week 9	Week 10	All	Not Completed
Stage 3: Time Series Analysis	Week 3	Week 9	Jouta	In Progress
Stage 4: Geospatial	Week 4	Week 13	Qianpin	In Progress
Interim Preparation
Gather feedback from Client	Week 4	Week 8	All	Not Completed
Prepare interim report and slides	Week 7	Week 8	Jouta	Completed
Application Building
Code the application	Week 6	Week 14	All	Not Completed
Testing the application	Week 10	Week 14	All	Not Completed
Gather feedback from Client	Week 12	Week 16	All	Not Completed
Iteration
Adjust analysis	Week 12	Week 13	Jouta	Not Completed
Refine results to improve clarity	Week 12	Week 13	Wan Theng, Qianpin	Not Completed
Final Preparation
Prepare Research Paper	Week 14	Week 14	All	Not Completed
Prepare Poster	Week 15	Week 15	Qianpin	Not Completed

Work Plan

• Fortnightly meetings with Supervisor
• Monthly meetings with Sponsor

Midterm Changes:

• Exploratory Analysis extended by 2 weeks due to delay caused by data complications
• Clustering module shifted to start after recess week
• Time Series was brought forward and will end earlier than initial deadline (Wk 9 instead of Wk 11)
• Geospatial extended by another 2 weeks due to sponsor data/online data compatability issues
• Web app changed to Rshiny app. Reasons being that Rshiny is interactive, free, and can accomodate a wider range of analytics. Extend by 2 weeks because we are new to this