@brian_plancher You mentioned a “TinyML Trash Car” which I think that is a great idea, it reminds me how my RC hacked tinyML car began.
I made vehicle “A” which had a servo motor to turn the front wheel that also was the drive wheel. A student did “B” for the end of year Robotics project fully 3D Printed from PLA with tires from flexible NijaFlex material. The car was WiFi driven and cell phone webpage controlled car.
I made “C” to be a WiFi websocket TensorflowJS machine learning car that had a cell phone attached to the front. Video here, but that started getting really complex deprecated github here, so we are now working on “D” using the Arduino Portenta with LoRa vision shield and edgeimpulse ML models since it reduces the complexity of the already complex situation.
Once we have our cars working at an acceptable level, I would really like to look back at the complete build idea. Your idea to forget how it looks just get it working is great. I wonder if you could get a grad student working on this?
1 Like
As we were talking about ChatGPT here is what I sent it: “Write a grad student grant application to design a tinyML Machine Learning car built from cheap electronics and trash as a recycling method of minimalistic education”
and here is it’s reply:
Dear Review Committee,
I am writing to request a grant for my project, “TinyML Machine Learning Car: A Recycling Method of Minimalistic Education”. My goal is to design and build a tiny machine learning car using cheap electronics and trash materials as a way to educate students on the principles of minimalism and recycling.
The tinyML Machine Learning Car will be built using inexpensive electronic components, such as a Raspberry Pi, a microcontroller, and a few sensors. The body of the car will be constructed using recycled materials, such as cardboard, plastic bottles, and old tires. The car will be able to navigate its surroundings using machine learning algorithms, which will be trained using a small dataset.
The project has several educational benefits. First, it will teach students the principles of minimalism, as they will have to design and build the car using a limited number of components and materials. Second, it will introduce them to the concept of recycling and encourage them to think about the environmental impact of their actions. Finally, it will provide a hands-on experience with machine learning, allowing them to see how it can be applied in practical situations.
I plan to use the grant funds to purchase the electronic components and materials needed to build the tinyML Machine Learning Car. I will also use the funds to develop educational resources, such as lesson plans and tutorials, to help students learn about the principles of minimalism, recycling, and machine learning.
I am confident that this project will provide a unique and engaging learning experience for students, and I look forward to the opportunity to bring it to life with your support. Thank you for considering my grant application.
Sincerely,
[Your Name]
1 Like
OK this is getting creepy impressive lol
And yes I love all of this – will try to get a student funded to do this! 
I am now getting 197 ms FOMO classifications using the $15 USD Seeed XIAO-esp32S3 and also trying to get a HobbyWing $21 USD brushed speed controller working, Along with a servo motor would dramatically reduce the base price for making some trash cars.