The X-5 Drone is an experimental concept for an autonomous aerial platform designed to complement the first human crews on the Martian surface. The main objective of X5 is to demonstrate the reliability and the capabilities of an autonomous solar aerial drone platform for surface surveys on Mars, by performing test demonstrations in analog sites on Earth. Thanks to its VTOL and automatic flying capabilities, the vehicle can autonomously achieve the mission objectives, performing take-off, hovering and landing maneuvers without the need for direct control from the crew. In one of its versions, the X5 can be powered by lightweight double-junction solar arrays, that can potentially assure a dawn-sunset flight autonomy.
The X5 payload is composed of two cameras (one fixed global-shutter camera and one for navigation) and a range of sensors, with the capability to host more mission-specific payloads.
The X-5 Drone has been tested by Crew 212 during the LATAM III analog mission at the Mars Desert Research Station (MDRS) in Utah, USA. The X5 MkII, with improved flight capabilities, will be tested during the AMADEE-20 analog mission that will take place in October 2020.
In 2018 I contibuted to the creation of DOME. The DOME Project (Drone Operations for Martian Environment) is a research group, built between universities and companies, which aims at developing new remote and autonomous operated aerial platforms to support operations on the martian surface. Aerial drone platforms have become a cutting-edge asset in a wide range of human operations such as medical, firefighting and military scenarios. In extraplanetary exploration, drones could bring both the high precision of surface rovers and the extensive area coverage capabilities of orbital spacecraft. Human exploration will need a reliable platform to cover large areas in a limited time, with the necessary precision and surface analysis capabilities.
In almost 60 years of exploration of Mars, humankind has tested a wide range of technologies to study its surface, using platforms such as rovers, probes and orbiters: Mars is the planet with the biggest robot population in the solar system.
Until now, less than the 1% of the martian surface has been explored in detail. New platforms are needed to rapidly increase our exploration capabilities if we want to seriously boost the first human missions to the Red Planet.
Aerial drone technology has considerably evolved in the past years, thanks to the increasing number of potential applications. Hardware miniaturization and deep-learning algorithms brought this technology to a fundamental role in high-risk scenarios. Mars, due to its geological and atmospheric properties, represents a totally new ground to expand the boundaries of this technology. Aerial drones can become a fundamental subsystem of human activities on Mars: logistics, safety inspections, search and rescue missions, and multispectral analysis can be safely left to swarms of autonomous flying drones.
The DOME project research background is focused on:
– On-field testing of aerial platforms to explore and support human activities
– Deep learning algorithms
– Solar power management
– Object tracking and obstacle avoidance
– Human-Machine Interaction
– Flight attitude in Mars atmosphere
– Remote sensing
– Aerial mapping
Neural networks and AI are an emerging field in the drone industry. Object recognition and tracking Algorithms are fundamental for real-time applications on Mars, since the huge communication delay. To efficiently fly on the martian surface we`ll require navigation software capable of detecting obstacles and updating the UAV route in real-time. Also, the same algorithms can be used to meet specific mission requirements such as outpost inspections, geological surveys and Search and Rescue operations. Without GPS the data coming from the onboard cameras are an important feature to achieve situational awareness and position.
My research effort in AI at DOME is focused on three main areas of computer vision:
– Object tracking
– Object recognition
– Pattern Recognition
The developed algorithms are implemented inside the drones onboard computers and are going through extensive field testing to ensure their reliability.