We had the incredible opportunity to participate in Hack the North, a prestigious student-led hackathon where students can spend a whole weekend designing innovative software and hardware projects.
In this blogpost, we’re shining a spotlight on one of the standout winners for the category “Best Use of Viam”, where Team Spyder achieved their ambitious goal to transform an old motorized Audi Spyder toy car into a mind-controlled marvel. They ingeniously integrated the Viam API with a Neurosity Crown, enabling them to trigger robot protocols using the power of their mind.
To get a glimpse of how this was made, we interviewed the creators of this project–keep reading for more!
Tell us about yourself and your team.
We are Team Spyder! We’re made up of four 3rd year engineering students at the University of Waterloo who are experienced in both engineering competitions and hackathons. This year, we set out to compete in Hack the North 2023 for fun, learning something new, and this year, a win!
Can you walk us through the initial concept of your build?
The initial idea struck when leftover electronic parts and a toy car faced the trash at the student center. Why not turn this into something?
Inspired by Michael Reeves' build, we fantasized about building a mind-controlled car, but aimed to make it even better.
Michael Reeve's mind-controlled car: the inspiration behind Spyder.
We took the car apart to see if our idea was even feasible–can we modify this to match what we need? Would we have the right electronics to make this happen?
Turns out when you have a bunch of overexcited engineering students working on a project like this one, it doesn’t matter. We’ll make our own electronics.
So, what does your build do?
We made several mechanical and electronic modifications to the car to meet our specifications, including:
- Designing and 3D printing a custom rack and pinion gear to adjust the steering system, coupled with an additional motor for automated steering.
- Upgrading the battery system for enhanced performance and durability.
- Incorporating advanced electronics such as a Raspberry Pi 4, buck converter, LED controller, custom power distribution board made by Ari, and motor driver.
Next was the software component. We used Viam to control the car electronics once they were all plugged in and soldered in.
For the mind-control aspect, we harnessed the Neurosity Crown and its AI, training it with a dataset derived from Patrick’s brain waves. This approach enabled the car to respond to Patrick's thoughts, moving forward when he intended it to.
Patrick using the Neurosity Crown to capture his brain signals.
We integrated the two software elements using scripts crafted by Melda, utilizing web sockets for connectivity.
Can you walk us through step-by-step on how you used Viam for your build?
We initially turned to Viam to validate our concept. Aaditya explored Viam’s documentation, employing a motor driver compatible with the Raspberry Pi. Following the guidelines provided, we connected it and leveraged the Viam server for system troubleshooting!
This step was crucial as it allowed us to confirm the integrity of our electrical system before writing any code. The Viam server was instrumental for our troubleshooting and sanity checks, saving us several hours.
It also enabled us to test our motors and mechanical systems to their limits. After validating our systems, we found the code generated by Viam and its SDKs straightforward to integrate into our codebase.
What were the most challenging aspects of your build?
One of the most challenging aspects was getting the right data from the Neurosity Crown. Since the electrodes sit on top of the scalp, it makes it difficult to capture clear brain waves.
Think about it like you’re listening to a conversation happening in a room but you’re outside the closed door. It’s muffled and sometimes you can’t quite get all the words.
Eventually, we got it to work after gathering a few datasets and Patrick’s thoughts were properly recognized!
What was the learning curve like for Viam's platform?
Viam’s platform and SDKs were very easy to pick up and use in our design process and codebase. Viam’s interface was simple and easy to use, but still packed with useful features. The SDKs were very straightforward and a quick read through the documentation was all it took to figure out how to effectively use them.
Any plans to upgrade your current hackathon build?
We’re considering integrating a webcam for Computer Vision, enabling the car to automatically slow down upon detecting humans in its path.
We also aim to enable simultaneous recognition of multiple commands, enhance our electromechanical systems to support a passenger in the car, and even introduce mind-controlled unlocking of the vehicle!
We have a few other ideas to use a Viam Rover for next year’s Hack the North project so stay tuned for that!
Any final thoughts or words of wisdom you would like to share with aspiring builders and innovators?
Don’t be afraid to try something new! We initially started making this project as a joke. Who would ever think about creating a mind-controlled children’s car?
We had a lot of fun making this project and learning how to work with neuro-technology and Viam. The win is definitely the cherry on top of the whole experience, but ultimately, we came out of it as better engineers and created fun memories to go with it.
Things won’t always go to plan. We had quite a few hiccups and had to pivot a few times, especially at the beginning, but it’s important that you try your best and go with it. Sometimes when things don’t work out initially, they end up becoming better.
Ready to create your own smart machine? Dive into our tutorials for immediate building, explore modular resources to enhance any project with Viam, and join our community for endless inspiration from innovative creations!"
