Aptiv - 2020
Vehicle Test Cases Identification during Real World Scenarios
Background
Aptiv PLC (NYSE: APTV) is a technology company that is transforming mobility with its portfolio of safe, green, and connected solutions for a diverse array of global customers. Aptiv will usher in the next generation of active safety, autonomous vehicles, smart cities, and connectivity by creating smart mobility solutions with a strategically focused data & power, central computing, and connected services portfolio. Aptiv addresses key mobility challenges through industry-leading software capabilities, software and systems architecture integration, and compute platforms that unlock next generation features - now. Headquartered in Gillingham, England, Aptiv has 147,000 employees and operates 14 technical centers, as well as manufacturing sites and customer support centers, in 45 countries.
Overall Goal
Design and develop a data mining algorithm to post-process vehicle logs from on-road Autonomous Driving testing in the US, Europe and Asia.
Project Goals
- Detail the scope, requirements, stakeholders, permissions, etc. for the system with the Aptiv team.
- Develop a functioning, scalable application that will automatically post-process vehicle logs from on-road driving.
- Integrate the application onto Aptiv’s data server in Southfield, MI.
- Create detailed product documentation for take-over by Aptiv team afterward, particularly for revision and updates.
- Document Project updates via source/version control.
Purpose of Algorithm
- Scan all vehicle logs for various environmental features that are not typically identified by an Autonomous Vehicle algorithm, such as overpasses, tunnels, bridges, road-side features, overhanging trees, types of vehicles and more. List provided by Aptiv, and expanded by MTH844 project team.
- Tag the vehicle log files corresponding to which features are in each file, for later querying.
Aptiv Input Deliverables
- Usernames and access to Aptiv’s network and systems.
- Requirements breakdown from Aptiv users.
- All necessary information on the data and the server to integrate to.
- A demonstration of a test vehicle (hardware, sensors, driving etc.).