This repository contains my submission for the Deloitte STEM Virtual Program. In this program, students are asked to help Daikibo Industrials (a global leader in manufacturing of heavy machinery) with a variety of problems that they are facing.
Daikibo is in the process of integrating IIoT (Industrial Internet-of-Things) devices to monitor, measure and analyze their manufacturing processes. Half of their infrastructure uses devices streaming telemetry data in one format, and the other half - in another. They need help to combine the two files.
By completing this task, I learned how to manipulate json files in Python as well as how to use unittest framework to write and run test for a python code.
Daikibo's tech team has converted all telemetry data collected from the 4 factories of the company:
- Daikibo Factory Meiyo (Tokyo, Japan)
- Daikibo Factory Seiko (Osaka, Japan)
- Daikibo Berlin (Berlin, Germany)
- Daikibo Shenzhen (Shenzhen, China)
Each location has 9 types of machines, sending a message every 10 min. Daikibo has been collecting this data for 1 month (May 2021) and they've just shared this data in the form of a single JSON file. The reason this client wanted to collect telemetry was to answer 2 questions:
- In which location did machines break the most?
- What are the machines that broke most often in that location?
By completing this task, I learned how to use the famous Analytics tool Tableau to build an interactive dashboard that helps the Daikibo's tech team to answer their questions.
Analyzing offline data is great, but having a real-time overview of processes and smart alerts when things break is even better. Daikibo's tech team would like us to build a real-time manufacturing status dashboard. The first task of this process is drafting a development proposal.
By completing this task, I learned how to draft a concise and convincing software development proposal.
A big news publication has revealed sensitive private information about Daikibo Industrials' β a production problem has caused their assembly lines to stop, threatening the smooth operation of supply chains relying on Daikiboβs products. The client suspects the security of their new status board may have been breached.
By completing this task, I learned how to read web log files, and how to identify spot suspicious requests. It turns out. For a little bit of context, the dashboard lives in Daikibo's Intranet. The only remote access to it would be through VPN tunneling. After navigating through the web log file, I noticed that the request of the user with the IP address of 192.168.0.101 starts off with a regular login and then browsing of the dashboard. But then it turns into a regular, once-per-hour automated check of the statuses in all 4 factories with no page resources being loaded and with obviously non-human punctuality.
After a worrisome number of internal complaints about gender inequality (in terms of pay), Daikibo Industrials wants us to help them investigate. The Forensic Tech team has built an algorithm to quantify the βlevel of gender pay equalityβ for most/all job roles within the company, in all company locations.
Doing this task allowed me to refresh my memory on Excel functions.