AI for Auto Prognostics

Project Overview 

The On-Board Diagnostics (OBD) project seeks to equip vehicles with an advanced data recording and analysis system using a Raspberry Pi microcontroller. This system will capture real-time performance metrics from vehicle OBD ports, storing data to inform vehicle diagnostics and maintenance decisions. Students involved in this project will work towards developing machine learning algorithms to monitor driving behaviors and vehicle performance, eventually laying the groundwork for predictive diagnostic capabilities that anticipate and alert drivers to potential issues before they occur.

Objectives 

  1. To successfully install and initialize an OBD data collection system capable of capturing real-time vehicle metrics.
  2. To develop and implement machine learning methods to analyze collected data for behavioral and vehicle health monitoring.
  3. To conceptualize and initiate a training protocol for an AI-based prognostics system, enabling predictive maintenance capabilities

Major Tasks 

The major tasks of the project are outlined below. These are nominal and are likely to change somewhat throughout the course of the year. These are meant to illustrate the general nature of the work that the position entails.

  1. Set up and test Raspberry Pi hardware and software integration with vehicle OBD systems.
  2. Conduct data collection trials, capturing driving behavior, engine metrics, and performance parameters.
  3. Develop robust data logging protocols and cloud-based storage solutions.
  4. Perform data preprocessing and analysis, preparing data sets for machine learning applications.
  5. Create, train, and validate machine learning models focused on behavior analysis and vehicle health prediction.
  6. Research and outline initial strategies for implementing predictive maintenance and prognostic systems.
  7. Evaluate the system’s usability and safety in real-world driving environments.

Preferred Skills and Interests 

  • Interest in automotive engineering and vehicle diagnostics
  • Programming skills (Python, Linux, Raspberry Pi familiarity preferred)
  • Enthusiasm for machine learning, AI, and predictive analytics
  • Comfort working with embedded systems and data acquisition
  • Analytical thinking and data processing skills
  • Commitment to hands-on, field-based testing and troubleshooting

How to Apply 

Applications will be reviewed by Dr. Gray and by continuing/former researchers from the team.  After a review of the application, our team will contact candidates to schedule a zoom screening interview.  Review for positions will be conducted either during Summer (for Fall on-boarding) or Winter (for Spring on-boarding). Please complete the MS Form, uploading the following deliverables; 

  • A brief (~1 page) essay or cover letter explaining which of the projects you are interested in, and why you think you might be a good fit for that project (or those projects).  If you are applying for multiple projects, extend your essay a little and describe your interest and qualifications for each position.  Be sure to let us know your major and where you are in your academic career (sophomore, junior, etc.) 
  • A resume outlining your work experience and education 

Please reach out to Dr. Gray (dagray3@vt.edu) if you have any questions or concerns. 

See All Projects
Student Responsibilities
Apply Today