One of India’s tallest accomplishments this decade is undoubtedly, the Mars Mission or Mangalyaan. India successfully sent a satellite to Mars on September 24, 2014, which has been in orbit since. This feat, spearheaded by Indian Space Research Organisation (ISRO) has inspired a generation of scientists, researchers and students to explore life beyond Earth with an array of new age technologies. Since then, there have been countless college-level competitions, globally and locally, aimed at nurturing and harnessing the potential of students to build robots and rovers meant for planetary exploration.

Manipal Institute of Technology, one of Karnataka’s foremost engineering colleges, has been grooming a group of students to build next generation rovers. This collective, named Mars Rover Manipal (MRM), has built a state-of-the-art Mars rover that has won them multiple accolades – a world rank of #7 at the System Acceptance Review stage of University Rover Challenge 2020, a world rank of 1 in the Indian Rover Design Challenge 2020, and the third rank in the International Mars Hackathon 2020. The Indian Rover Challenge, is an annual robotics competition for students to build next-generation Mars rovers. The University Rover Challenge, organized by the Mars Society, takes place in the Mars Desert Research Station, Utah, USA – the site where NASA tests their Mars rovers before launch.

A Multidisciplinary Approach To Building The Rover:

Mars Rover Manipal functions primarily as a Mars rover development team and a robotics research organization. Owing to the multi-disciplinary nature of robotics and multi-functional attributes that make a rover effective, the team aims work at the boundaries and intersections of traditional sciences and engineering disciplines such as mechanical engineering, electronics, software development, and biotechnology. To achieve this goal, the team is divided into six subsystems – Artificial Intelligence, Electronics, Mechanical, Science, Research, and Management.

The Mechanical subsystem designs the rover and manufactures it in-house at the workshop in MIT, Karnataka. The Electronics subsystem is responsible for making the rover functional by designing PCBs, Motor Drivers, and connecting the hardware required to power an autonomous Mars Rover. The AI subsystem gives the rover life. They create programs that result in autonomous traversal and task handling by utilizing complex machine learning systems and algorithms to perceive the environment of the rover and take decisions. The Science subsystem handles the biological aspects of the competitions and is responsible for powering the rover’s search for extra-terrestrial life. Their chemical tests and other tests for bio-signatures are then implemented on the rover by the other subsystems to fulfil the primary function of the rover as an onboard, autonomous scientific laboratory.

The Research subsystem works on the cutting edge of Robotics and Artificial Intelligence on problem statements that may or may not be directly related to the Mars rover. They have presented their findings in research papers at the national level and international conferences. The Management subsystem is responsible for the overall functioning of the team, ensuring that they have an adequate budget by approaching companies for sponsorship, handling permission, and logistics involved in issues like shipping, and ensuring that the work is always running on time. They also look after the social media and website of the team to ensure that their work can reach far and wide.

(The Mars Rover Manipal design being tested at the MIT campus, Manipal, Karnataka)

How AI brings Mars Rover Manipal’s rover to life:

The AI sub-system is crucial to the success of this project. It deals with processing and interpreting data from various sensors; enabling the rover to traverse autonomously without human interference. The subsystem also deals with designing deep learning architectures to solve various state-of-the-art problems encountered during the development of rover.

Continuing its endeavour to support cutting edge research in AI, GPU & visual computing leader NVIDIA has provided edge system Jetson TX2 to MRM. “The autonomous task requires the rover to run state-of-the-art neural networks, for vision tasks. These tasks require heavy computation and the NVIDIA Jetson TX2 is the most effective computer on the edge for this in the market,” says the team. Jetson TX2’s blend of low power consumption and highly powerful computation helped the team in competitions and research work, allowing them to train complex deep learning models and deploy them along with powerful image and point cloud processing capabilities.

In addition, NVIDIA’s Jetson TX2 takes inputs from several onboard sensors like LIDAR, stereo cameras, GPS and IMU. It's also used for an autonomous robotic arm that takes in commands in the English language to generate feasible paths to follow, by processing data about the environment from various stereo cameras. In addition to this, the rover makes use of many innovative mechanisms such as a carbon fibre space frame chassis and a first-of-its-kind custom five bar suspension to keep the chassis and internal hardware stable.

There is wide use of 3D printed parts for the small, critical components. ARM cortex microcontrollers are used to take commands from the autonomous systems to provide lower-level commands to the rover through custom-designed PCBs, in essence acting as a nervous system for the robot. The team also claims to have a science module with an innovative pumping mechanism on which the rover conducts chemical and biochemical tests to test rocks and soil for signs of life.

With the purpose of providing an intelligent robot that is primed to survive the toughest conditions present on the red planet, search for signs of life, support existing life in Mars and eventually facilitate interaction with other humans (if astronauts do reach Mars one day); Mars Rover Manipal’s next goal is to achieve a podium finish at the University Rover Challenge as well as ramp up research efforts in robotics, AI, electronics, astrobiology, and biotechnology.