During the mission of Vikram Lander approaching the surface of the moon near the south pole, the AI & ML used in the Lander has come into work. At the time of lander is in 150 meters altitude above the surface of the moon an obstacle detection camera has identified an obstacle right beneath the area where the lander is supposed to be landed. As soon as the identification, the lander moved a few steps towards the right and made it successful soft landing on the surface of the moon.
What kind of AI & ML is used in order to move the lander on the identification of an obstacle?
They have installed an object detection camera right under the Vikram lander, which will identify the obstacles.
During the development phase of the Vikram lander, ISRO fed the data and trained the Vikram lander with various obstacles that might be present on the surface of the moon, like sand, stones, and craters.
![Surface of moon]()
In Physics, there is a concept of finding the size of an object using its shadows. The same principle is trained in the AI module present in the Vikram Lander.
So, what it does is that the camera identifies the obstacle and sends that image to the AI present in the lander.
Then the AI Module measures the shadow of the obstacle and calculates the size of the object that is present. Once the size of the object is identified, it then calculates the number of steps it must move in order to avoid contact with the obstacle.
Then the AI module moves the lander a few steps so that the lander doesn’t make contact with the obstacle.
This is how one of the AI & ML had been used in the Lander.