Week 8: Refining the CNN & Finalizing Hardware Setup

Anshul B -

Hello and welcome back to my blog! My name is Anshul, and this week, we have been focusing on refining our software, troubleshooting hardware connections, and preparing for the final integration of our system. As we approach week 9, we want to make sure every component is functioning as expected.

Solving the components…

My primary focus this week has been refining the CNN model. I successfully trained the model to accurately predict objects. Also, I have helped my teammate in developing the Depth-from-Defocus (DFD) algorithm. However, we are encountering issues with the OpenCV library in Python, which is causing difficulties in connecting to our camera. Another hardware problem of ours is being solved by working on soldering and properly wiring the Raspberry Pi to the motor driver.

A key accomplishment this week is that we have officially started implementing our code into the GitHub pipeline. I successfully uploaded the completed CNN model. Meanwhile, my teammate has been developing the DFD algorithm and steering commands. Last week, we encountered issues with the car’s movement using the Bluetooth mouse, particularly with turning. This week, we are addressing those problems to improve control and responsiveness.

Progress Toward Final Testing

Our main goal this week has been finalizing all hardware and software components. With the CNN model in place, the DFD algorithm/steering commands in progress, and the motor wiring being refined, we are preparing to integrate all components. Our goal is to have a fully functional system ready for testing next week. Also, I have attached a picture of our final poster for the symposium. Thank you for taking your time to read my blog, and stay tuned for updates!

More Posts

Comments:

All viewpoints are welcome but profane, threatening, disrespectful, or harassing comments will not be tolerated and are subject to moderation up to, and including, full deletion.

    shriya_s
    Hi Anshul, this is a very fascinating and exciting project! I'm glad to see that you are making so much progress. In the poster for the symposium, it is written that you plan to involve ICON in food-delivery. What do you think are some challenges associated with that, which may prevent you from developing the app right now (e.g., time, resources, bugs, etc.)?
    anshul_b
    Hey Shriya, that's an amazing question. Our biggest challenge towards building the app is getting the car to move and perform object avoidance. Then, we have other phases and will eventually get to the app.
    camille_bennett
    Hi Anshul, great to hear you are looking at testing the whole system. Any lessons you are taking away from this experience?
    Anshul Baddi
    Hey Ms. Bennett, The biggest lesson I'm taking away from this experience is that coding a self-driving car was a lot harder than I expected.

Leave a Reply

Your email address will not be published. Required fields are marked *