Mars on Earth

Aryan Jha
4 min readDec 12, 2020

--

Pretty much every kid at one point or another dreamed to be an astronaut. Some actually become one. But nothing can fully prepare them for life on another planet. Maybe some semi-realistic simulations can help. But nothing can truly make them ready. Until now.

Astronaut in space

Imagine if you took a sample of Mars, and put it in a room. Change a few properties here and there, and now you have a Mars simulation. We can’t do that, unfortunately, but we have the next best thing: A perfect clone of Mars. That’s right. Using the power of quantum computers and AI, we can do this. Why do we need a Mars simulation? Putting an astronaut on Mars without training is like asking a baby to get a job. They wouldn’t know how to do anything. It is our job to make their training as in-depth and realistic as possible, so they can succeed. There is training for astronauts, but it isn’t completely realistic. We plan to change that with our simulation. But how would it work?

At first, it may look very hard, maybe impossible. But, breaking it down into smaller parts will undoubtedly help. There are 2 parts: collecting samples and building the simulation. I will explain each part in detail.

Part 1: The sample collection

Mars rovers are hardly a new thing. They have been there for years, collecting information. What if we can harvest their power to do something bigger, such as simulating Mars on Earth?

NASA Scientists Working on Robot

It’s possible. Using AI, rovers can find the best places to take samples. The rovers will be trained to collect many samples, but only send the most high quality ones to Earth. Using convolution neural networks (CNN for short), the rovers can recognize images. CNNs can recognize different objects in a picture and assign them different levels of importance. This is very useful. For example, it could be used to reject a sample because it recognizes the structure and knows that many of those samples have already been sent back.

Recurrent neural networks (RNN) can describe what is in the pictures, which makes it much easier for everyone on Earth. The samples are crucial to the simulation part, as they are what is being replicated. This is actually already possible. We don’t need any groundbreaking discoveries yet. Rovers are already analyzing samples and sending their discoveries to Earth. The real fun comes in the next part.

Mars

Part 2: The simulation

This is where our idea really starts to shine. The Mars sample data, after being sent to Earth, is then recreated. By using materials similar to those found on Mars, we can make a full environment. This could help prepare astronauts for anything. VR and haptic suits will be able to simulate actually being there, instead of being in a recreation. To understand this, I first have to explain how haptic suits work.

Haptic suits simulate touch using vibrations and electric pulses. Basically, it’s VR but with touch. How could this help? Haptic suits and VR can simulate wind and pressure on you. The VR is simulating the visual, while the haptic suit is simulating the feel of it. Why can’t we just do this in real life, instead of using VR? Simulating Mars’ famous dust storms on Earth would be very hard. VR and haptic suits can make everything much easier by doing it in a virtual world.

All of this is nice, but how are quantum computers involved? Well, making a realistic Mars environment would be really annoying to make by hand. Quantum computers in robots can visualize the bigger picture and work together to place everything exactly like it is on Mars. This can save a lot of time and energy, as well as not allowing as many errors.

Now, we have a hyper realistic Mars simulation. Using this technology, we can establish a permanent presence on Mars. Think about this new world, and what discoveries we can make. Think about the fact that it is only 1 product away.

--

--

Aryan Jha
Aryan Jha

Written by Aryan Jha

Student interested in Reinforcement Learning.

No responses yet