Robot Chef
Intro
My internship at SpaceX Starlink (Summer of 2024) was very fun, I learned a lot of technical things and life skills! One of the things I learned was how cooking food for yourself is extremely time consuming.
Problem/Inspiration
I used to eat 3 meals a day. At SpaceX, I ate 2. Why? Because as a working adult trying to build the future of humanity, there is simply not enough time in the day for me to cook and eat three meals while still getting work done....so let's build a robot that can cook me food!
The Process
The plan was....actually not that simple this time:
1. Narrow my scope. Building a general robot that can do anything right from the get-go was impossible.
2. Design it in CAD.
3. Build the robot arm.
4. Design the electrical system.
5. Build the electrical system.
6. Think about the code for the algorithm, AI, computer vision, CNNs, etc.
7. Cry myself to sleep about the code.
It just so happened that at the time of the build, there was a Hackster contest going on sponsored by AMD, so I decided to enter the competition cause why not? Here is the Hackster post. Check that out if you want to see pictures, detailed step-by-step process, and my story.
Remarks
I am extremely proud of the cable-actuated wrist system and the use of salvaged motors for the arm control, printer rods for the robot arm axes, and potentiometers for feedback. Even though I was super creative, it was probably the worst thing I could have done, since I spent far too much time trying to optimize fricion and backlash, trying to debug why the position accuracy would start to drift every time I operated the arm (because of slipping), and re-printing the PLA parts that fall out of tolerance break due to the motors getting too warm. On top of that, I had to figure out how to simulatenously use a new piece of hardware and a new software stack. Ahhh! This project was by far the most complex and stressful thing I've pulled off and probably not too good for my health in the long run.
But would I do it again?
Of course, because I'm insane.
Also, I am most definitely going to build a newer and even more insane version of this, so keep your eyes peeled!
My internship at SpaceX Starlink (Summer of 2024) was very fun, I learned a lot of technical things and life skills! One of the things I learned was how cooking food for yourself is extremely time consuming.
Problem/Inspiration
I used to eat 3 meals a day. At SpaceX, I ate 2. Why? Because as a working adult trying to build the future of humanity, there is simply not enough time in the day for me to cook and eat three meals while still getting work done....so let's build a robot that can cook me food!
The Process
The plan was....actually not that simple this time:
1. Narrow my scope. Building a general robot that can do anything right from the get-go was impossible.
2. Design it in CAD.
3. Build the robot arm.
4. Design the electrical system.
5. Build the electrical system.
6. Think about the code for the algorithm, AI, computer vision, CNNs, etc.
7. Cry myself to sleep about the code.
It just so happened that at the time of the build, there was a Hackster contest going on sponsored by AMD, so I decided to enter the competition cause why not? Here is the Hackster post. Check that out if you want to see pictures, detailed step-by-step process, and my story.
Remarks
I am extremely proud of the cable-actuated wrist system and the use of salvaged motors for the arm control, printer rods for the robot arm axes, and potentiometers for feedback. Even though I was super creative, it was probably the worst thing I could have done, since I spent far too much time trying to optimize fricion and backlash, trying to debug why the position accuracy would start to drift every time I operated the arm (because of slipping), and re-printing the PLA parts that fall out of tolerance break due to the motors getting too warm. On top of that, I had to figure out how to simulatenously use a new piece of hardware and a new software stack. Ahhh! This project was by far the most complex and stressful thing I've pulled off and probably not too good for my health in the long run.
But would I do it again?
Of course, because I'm insane.
Also, I am most definitely going to build a newer and even more insane version of this, so keep your eyes peeled!