A Novel Location and Attitude Determination System for a Swarm of High-Speed Martian Tumbleweed Rovers
- Paper ID
92470
- author
- company
Team Tumbleweed; International Space University (ISU)
- country
United Kingdom
- year
2025
- abstract
Traditional Mars rovers have covered limited terrain due to their slow speeds. While ESA targets speeds of 1 m/s for next-generation rovers for the moon, which has few obstacles compared to Mars, our Team Tumbleweed swarm of rovers can achieve speeds of 5 m/s, through an innovative wind-driven, rolling, lightweight design. This increased mobility and reduced weight enables swarm operations that will collect much more data at a fraction of the cost compared to traditional single-rover missions. Accurate Location and Attitude Determination (LAD) is crucial for fast-moving rovers to properly geotag scientific measurements. The absence of conventional GNSS on Mars is an issue which we try to address using a novel LAD system that uses miniaturized low-complexity, low-mass Commercial Off-The-Shelf (COTS) sensors for Tumbleweed's unique mass and power constraints. The system combines IMUs and rotary sensors during movement with feature detection and satellite navigation signals being used during strategic stops, triggered when position error exceeds defined thresholds. During the Martian night when the rovers operate only on battery, star sensors and satellite navigation signals are used to correct the location error. The study begins with a thorough literature survey of prevalent LAD sensors. The sensors are classified according to mass, size and processing power based on objective criteria. The sensors are reviewed, highlighting their strengths and weaknesses. The findings guide the sensor selection and placement for the swarm of rovers. The impact of stopping and starting to correct errors is shown with initial results of sensor fusion and noise filtering. We developed and tested our LAD system using Unity, a real-time simulation platform to model Tumbleweed rover movement and sensor outputs. The simulation incorporates sensor error and noise profiles based on COTS datasheets, validated through our product testing. We integrated high-resolution Martian elevation data and wind profiles to create an accurate representation of Mars conditions. This simulation environment is available to other teams for mission planning, trajectory estimation, and mechanical design testing.