Designers use randomness to create things like cars, buildings and bridges. They use randomness for common engineering problems, like strength, material costs and weight. Fortunately, computers are here to the rescue and can help us solve a lot of these issues. One of the techniques computers can use to solve these issues is Machine Learning.
To demonstrate, we’re going to create our own virtual creature, and show how machine learning can help it take some first baby steps!
Machine Learning: Design a creature and teach it to walk!
How to teach with machine learning?
First of all, what even is machine learning? Machine learning is simply the process of teaching a computer how to do something by giving the computer lots of data to process. In the activity below, we want to teach a virtual creature to walk. But how? Well, the same way we humans learn to walk!
You and I learn how to walk through trial and error when we are infants. But how do we know what walking actually is and how do we know we’re doing it right? Well…we look at examples of other humans around us walking, and Mommy & Daddy are excited when we can cross the room. So, we look for examples of how to do something and then figure out if we are on the right track. We can tell a computer how to do the exact same thing. This is machine learning.
Learning through trial and error
To make our creature, we first need to give it some joints, bones and muscles so that it can have a way of getting around. Then we use some fancy algorithms to teach the creature by getting it to try and walk over and over again.
Activity: Now let’s evolve our creature
Now it’s time to start designing your creature! You can create your creature and teach it how to walk by clicking the “Play“ button:
Randomness is also used in automotive design so, if cars are your thing, check out these links & videos about kids and car design!
Design Your Own Car: STEM Online Learning
How are Cars Made?
Kids Describe Cars of the Future